You're viewing the old website

Free Energy is all about freedom:
Power to the people -- literally and figuratively

 "Free Energy" 

News XML
- PESN Specials
- About
- Pure Energy Blog
- Daily FE News
- Features
- Free Energy Now
- This Week in FE
- Newsletter
- How you can help
- Submit  
- Subscribe


Energy Topics

• Alt Fuels
• Anti-Gravity
• Batteries
• Betavoltaic
• Biofuels
 - BioDiesel
 - BioElectricity
 - Biomass
• Body Electric
• Brown's Gas
• Cold Fusion
• Conservation
• Electrolysis
• Electromagnetic OU
• Fuel Cells
• Fuel Efficiency
 - Electric Vehicles
 - Engines
 - Hydroxy
• Fusion
• Geothermal
• Gravity Motors
• Human Powered
• Hydro
• Hydrogen
• Joe Cells
• Lighting
• Magnet Motors
• Nanotechnology
• Nuclear
• Nucl. Remediation
• Oil
• Piezoelectric
• Plasma
• River
• Salt Water Mix
• Solar
• Solid State Gen.
• Tesla Turbines
• Thermal Electric
• Tidal
• Vortex
• Waste to Energy
• Water
 - Water as Fuel
• Wave
• Wind
• Wireless Electricity
• Zero Point Energy
• MORE . . .

Open Source
• Freddy's Cell
• Bedini SG
• Safe Haven Villages
• MORE . . .

• Awards
• Conservation
• Conspiracy
• Directories
• Investment
• Kudos
• Legal
• Organizations
• Plastic and Energy
• Recycling
• Suppression
• Tools
• Trends
• MORE . . .

• OverUnity Forum
• Employment
• Events
• Humor
• Magazines
• Movies
• Newsletters
• Discuss. Groups

• Store
• Buyer Beware
- - - - - - - - - -
- Donate
- Contact




You are here: > News > March 31, 2014

Sterling's first H-Cat calorimetric test points to anomalous heat

A report of Sterling Allan's test of the H-Cat (HHO through catalytic converter) in water bath at Frank Crowther's house, March 25-27, 2014; achieving around 77% input/output efficiency. Conventional electrolysis and fuel cell limits combined give a 64% efficiency maximum, pointing to an anomalous effect.

Draft essentially completed: April 1, 4:30 am Mountain (you can pass on the April Fools humor, thanks). I just have some final proofreading.

Table of Contents:
- Experiment Introduction 
- Definition: HHO Gas

- Methods, Design, Materials
 o Sterling's H-Cat Heater Design (March 15)
 o Sterling's Calorimetry Protocol (March 22)

- Photos
- Results: Videos
 o Live Recording: Set-up
 o Live Recording: First Half of Test
 o Live Recording: Last Half of Test
 o 1 of 8: Setting Up
 o 2 of 8: Adding Electrolyte
 o 3 of 8: First HHO Flow-Rate Test
 o 4 of 8: Explanation to Kids
 o 5 of 8: First Data Reading
 o 6 of 8: Data Reading Protocol
 o 7 of 8: Evening Observations
 o 8 of 8: This Week in Free Energy REPORT 
- Results: Data (XLS)
- Calculations
- Corrections

- Discussion
- Comparison to Other Electric Heat
- Conclusions

- Next Time: Suggestions
- Design Proposal: Basic Wind Tunnel 
- Design Proposal: Heat Exchanger

- In Other News
- Donate

By Sterling D. Allan  
Pure Energy Systems News

Well, we did it. We were the first to run a test of Justin Church's H-Cat in a water bath to carefully measure the input versus output energy according to the procedure I published March 22, starting on March 26, and finishing the next day. This is in preparation for the H-Cat Heater open source project I launched.

The test took place at Frank Crowther's house in Ephraim, Utah, USA -- about a five minute drive away from my house. Frank's house is also where I first experienced Brown's gas a decade ago, using an ER 1200 by George Wiseman. That's when I got excited about Brown's gas in the first place, especially when we learned by taking a sample to a nearby gentleman who is able to do elemental analysis, that our HHO gas sample contained tritium!


For three weeks now, we've been giving a lot of attention to the concept presented by Justin Church of what he calls the H-Cat, which takes HHO gas (highly energized H) and runs it across the nano-platinum and nano-palladium contained in off-the-shelf catalytic converters (yes, the same found in most automobiles for the last decade or so), resulting in copious amounts of heat -- likely because of an LENR (aka cold fusion) process.

The purpose of the present test was to put the catalytic converter in a water bath to measure the energy output in the form of calories or joules transferred to the water, and compare that to the input energy going to the HHO generator in the form of Volts x Amps = Watts over time, converted to joules. We were hoping to see some significant input:output efficiency indicating that there is more than just catalysis of HHO gas back to H2O, suggesting that cold fusion or some other anomalous reaction is taking place -- harnessing the wheelwork of nature.

Even though our efficiency was in the range of 77%, it seems to me that we did achieve a demonstration of an anomalous heat effect, considering that a combination of comparable electrolysis and fuel cell effects would predict a peak efficiency of 60%; and also considering that only about 21% of the input HHO gas ended up as water in the bottom of the catalytic converter. Something else is going on.

Definition: HHO Gas

When I say "HHO gas" I am referring to the product of common-ducted (not separated into H2 and O2) gas emerging from the electrolysis of water. It is also called Brown's gas, Rhodes gas, hydroxy, oxyhydrogen. I prefer "HHO" both for its brevity but also because it points to a molecular form, which I think could be an energized, linear form of water that is gaseous, and discharges its power with great energy.

HHO gas is not well understood yet. There are many mysteries about its characteristics that point to some great possibilities. Some people claim to be able to run vehicles on water using HHO gas. Others improve mileage by adding HHO to the fuel intake. It's used for welding dissimilar metals; and even for neutralizing radioactive elements.

Bear in mind, too, that not all HHO gas is created equal. There are probably potentially as many varieties of HHO generator machines as there are varieties of dogs. Part of optimizing this effect will be exploring which of these systems work the best in this application.


We spent about three hours Tuesday (March 25) night preparing, then another seven hours on Wednesday getting set up.

Experimental Design

The design of the experiment was according to these two documents:
 o Sterling's H-Cat Heater Design (March 15)
 o Sterling's Calorimetry Protocol (March 22)


HHO Generator Apparatus

We used the 31-plate HHO generator by Steve D at We also used the 12 Liter per Minute Flash Arrestor, Heavy-Duty HHO Bubbler/Scrubber, and the 3 Quart Reservoir they sell. Steve has been experimenting with the H-Cat as well, so he was very helpful in guiding us in our selection and implementation. We placed half a dozen calls to him while setting up and running the experiment.

At first, when hooking up the tubes, I was putting a hose clamp on them as well, but we stopped doing that on the latter portion. Steve said it is recommended for permanent settings, but not necessary for temporary set-ups.

We added 3 liters of distilled water to the reservoir, with 22.5 teaspoons of KOH crystal total. (Frank has a water distiller in the same external room where we were running the experiment.)

At first, we had hooked up the tubing wrong. We had the outgoing HHO gas going directly to the heavy-duty bubbler, instead of going first to the reservoir, so we had problems with the electrolyzer water running into the heavy-duty bubbler. We rinsed out the extra bubbler and hooked the the tubing properly: first to the reservoir (which doubles as a bubbler), then to the extra bubbler. We had two flashback arresters in the tubing between the extra bubbler and the catalytic converter.

We filled the extra bubbler to 2/3 full with regular tap water.

Power Supply

Rather than buying a new power supply, we used a Ham radio power supply from Frank's neighbor. It was adequate for our test, but for more accuracy, it would be nice if we had regulation on the volts and the amps so they could be selected and held constant. It would also be nice if we had a power supply that had an amp and volt meter built in.

The power supply we were using was an "Adjust-A-Built", 0-40 Volt, by Standard Electric Prod. Co. of Dayton, OH. I can't find a relevant link for them or for the product online. It is quite old.

Water Basin and Insulation

We used a Behrens 13.75 gallon circular, metal tub (hot dipped galvanized steel; shipping weight 4.25 lbs), and filled it with 46 liters of water. I did this by filling a 2-liter bottle on which Frank had accurately measured 2 liters and put a mark on the bottle at that point. I used a sink faucet and targeted 55 °F [Correction: was actually 65 °F] which was [approximately] the ambient temperature likely out in the greenhouse Frank has exterior to his dome home. The day was overcast, so we were expecting the ambient temperature to hold fairly steady, which it did. The ambient temperature was 52.5 °F [Correction: was actually probably 62.5 °F or lower] in the morning when I filled the tub, but had reached 55 °F [Correction: was actually 65 °F] by the time we began the experiment.

We set the tub on 1/2-inch Styrofoam, then wrapped the perimeter of the tub with R-19 insulation, which was about 4 inches higher than the tub. On top of that, we placed cardboard, on which we placed a foam couch cushion; and we placed a scrap piece of R-19 insulation over the opening where we would be taking temperature readings.

I modified this at 8:43 pm, pouring packing peanuts on top of the water so that it had 4-6 inches of these. When we did the cool-down portion of the test after turning off the HHO generator, the temperature dropped only 0.5 ΊC every hour for six hours. It continued to cool down at that rate after I removed the packing peanuts, for the final four hours of the cool-down measurements, but the detla T to ambient wasn't as great then, even though was snowing by then, outside. The ambient had dropped to 44 °F during the first six hours of the cool-down test, and was up to 49 °F during the last four hours. 

Catalytic Converter

We got our catalytic converter from who were excellent to work with. They took an off-the shelf metal matrix embedded with nano-palladium and nano-rhodium and put it inside an off-the-shelf 4" (outer diameter) stainless steel tube. The matrix with 300 cells per inch was 4" long. On the bottom of the matrix, the stainless steel tube tapered down to a 2" diameter, that was 2.5 inches long at 2" diameter. 

Insulating the Top of the Catalytic Converter

With the metal matrix pushed down against where the shaft tapered, there was about 2.5 inches of empty shaft at the top, which we filled with four pieces of 1/4 inch ceramic insulation (total weight: 1.25 oz) cut to a 4 inch diameter so it would have a snug fit inside the tube. This was to both diffuse the incoming HHO gas so it would hit the entire matrix opening evening, as well as to insulate the heat of the matrix from the incoming concentrated HHO gas to prevent backflash (we did not experience backflash).

Capping the Catalytic Converter

We covered the openings of the catalytic converter plumbing caps that were about 1/4 inch smaller in diameter than the opening, so the stretching required to get them around the openings was substantial. In the rubber plumbing caps, Frank drilled a hole that was about 1/8" smaller than the diameter of the tubing, so there was a strong pressure fit when the tubing was twisted through it. We topped that off with silicon on the inside, which we let cure overnight. When we put the caps on the catalytic converter, we also put the metal tighteners (larger version of the hose clamp) on to further secure it (probably unnecessary in our temporary set-up).

Two days later, after the test, when we took the caps off the catalytic converter, one of the silicon seals had come undone, but the pressure fit seemed to be adequate. We put water inside the caps and wiggled the tube, but no water escaped. The air-tight seal seemed to be complete.

We added the capped catalytic converter to the water bath at around 11:15 am, more than five hours before commencing the experiment. At that point, we measured both the ambient temperature and the water bath to be at 55 °F  [Correction: was probably closer to 65 °F]. We used electrical tape to strap half a brick (weight: 1 lb 14 oz) to the bottom of the catalytic converter so it would not float on the top of the water. We also added another 4 liters of water (making the total 46 L) to better cover the top end of the catalytic converter, which was still protruding above the top of the water about 1/4 of an inch at this point. 

The end of the exhaust tube from the bottom of the catalytic converter was kept out of the water, and was strapped to one of the tub handles with a rubber band to hold it in place.

Volt and Amp Meters

We used my UNI-T UT60A RS232C multimeter to measure the volts, connecting across the positive and negative leads from the power supply.

At first, we were using that meter to measure the amps, but because we surpassed the 10 amp limit of the meter, that function stopped working. It took us 2.5 hours to hunt down a clamp-on amp meter, which we borrowed from electrician Josh Peters (very reluctantly) of Custom Electric (thanks, Josh). I returned it to him the next morning, getting my collateral back.

[It turns out my dad, who lives a 1/2 hour drive away, has a clamp-on DC amp meter that we can borrow next time.]


See Sterling's Calorimetry Protocol (March 22)

Frank connected up the power cables from the power supply terminals to the HHO terminals, according to the instructions Steve D provided for the 31-plate cell.

To briefly summarize the procedure for running the system..., we ran Steve's 31-cell HHO generator on around 10.6 Volts and between 10.7 and 14.6 amps, using a concentration of 7.5 tablespoons of KOH per liter of distilled water. The HHO gas produced went up through the reservoir, which served as the first bubbler (to prevent backflash to the HHO generator). From there it went to a second bubbler, followed by two flashback arrestors, before venting into a catalytic converter with metal substrate embedded with nano-palladium and nano-rhodium. Any water formed would accumulate in the bottom of the catalytic converter's stainless steel tube and protruding hose. Any residual gasses would escape by bubbling through that water and venting into the surroundings.

After the first hour of running the test, we made sure to stir the water by moving a long, flat (1.5" x 1/4" x 5') wooden stick back and forth for 30-60 seconds before taking the water temperature reading.

We did our curiosity gas flow rate measurements by setting the power level, filling the 2L bottle with water, inverting it into a 5 gallon bucket of water, then running the gas into the bottle until the gas hit the "2L" mark that Frank had drawn on the bottle, and noting the time difference between "start" and "full". I say "curiosity," because the gas flow rate was not necessary for the input:output equation. It was just for a point of interest. (It turns out, though, that this secondary measurement helped us calibrate the system and qualify our amp readings as nominally accurate; so it is very good that we took these measurements.)

At first, we had a tube connected via a "T" to a U tube for measuring the pressure of the gas going through the tube from the first bubbler. It had some water in the bottom of the U, by which we could calculate the pressure based on how far up the other side of the U it went. However, Darrell Jacobson, who was helping us, pointed out that the pressure required to get the HHO gas through the flashback arrestors was enough to blow that water out the end of the U and create a void in the tube for the HHO gas to escape; so we scrapped that idea.


We never experienced a backflash during operation. There was always someone present during the first 8 hours of the experiment.

The temperature of the reservoir (108 to 110 °F), with the HHO gas bubbling through it, was substantially higher than the highest temperature of the HHO generator (97 °F). Steve D said that in his experience, "the reservoir is always warmer than the generator." At least 6 hours after I turned the power off to the HHO generator, when everything else (but the water bath) had equilibrated at ambient temperature, the reservoir was still about 6 °F above ambient, but that was probably just due to the caloric retention of that volume of water. Six hours after that, it was at ambient, like everything else.

The temperature of the extra bubbler was usually the same as ambient, sometimes a degree F cooler.

The back, top end of the power supply was hot, between 135 and 145 °F. Before the experiment, when we were testing flow rates, the power supply began smoking when we had the HHO generator producing 3 L/minute, so we turned it back down into the region of 1 L/minute.

We didn't yet have an amp meter when we were testing different power levels and rates of gas production until the last reading. Here is what we measured, when testing different power settings starting around 2:50 pm, prior to commencing the catalytic converter test.

  • 10.79 Volts: 2 L in 140 sec = 0.86 L/minute
  • 12.5 Volts: 2 L in 44 sec = 2.73 L/minute (power supply started smoking)
  • 11.04 Volts: 2 L in 100 sec = 1.20 L/minute
  • 10.73 Volts: 2 L in 138 sec = 0.87 L/minute

That last measurement was taken at 4:44 pm, just before we started the experiment at 4:48 pm. It's not needed for the calculations. When I talked to Steve that evening, giving him a rough average of amps and volts (forgetting we had recorded the volts and amps for the 4:44 test we did just before starting the catalytic test), Steve estimated, based on the volts and amps I read to him, that it was around 0.8 L/minute. When I emailed him the exact volts and amps we measured at 4:44 pm for the last flow rate test: 10.73 V, 11.55 A, he replied that his algorithm predicted a flow rate of between 700 and 750 mlpm -- lower than what we measured, suggesting, by way of calibration, that if anything, the amp reading was slightly (~15-20%) higher than actual, so the numbers we recorded were on the conservative end, rather than the other way around. A higher amp number will give a larger Watt input reading, which will reduce the calculated "efficiency".

Starting at 3:39 pm, we ran the HHO generator continuously at a setting that started at 10.85 V. By the time the experiment started at 4:48 pm, the potential had dropped to 10.73 Volts. And it continued to drop to 10.55 V at 5:29 pm, which was the lowest it ever got.

We left the power supply control knob at the same setting for the duration of the experiment.

At 12:22 pm, when we first used the glass thermometer, the water temperature was 18.5 °C. By the time we began the H-Cat experiment at 4:48 pm, the water temperature had dropped to 18.0 °C, while the ambient temperature went from 52.5 °F to 55 °F [Correction: was actually probably more like 62.5 to 65 °F].

Darrell Jacobson brought the glass thermometer with him. It was about 1 foot in length and was separated into 10 degree segments about every inch, with one small line for each degree, and a larger line for each 10 degrees. We could read it with an accuracy of about +/- 2.5 °C.

The comments about this make it sounds like this was a fatal assumption, nullifying the entire test. In particular, Asterix said: "An AC clamp-on probe is basically a transformer--it doesn't measure anything but the AC component of a current--the DC component is invisible to it. There are AC/DC current probes, but they look quite different from what you used-in particular, because they're active devices, usually using Hall-effect sensors to measure the magnetic field produced by a DC current. Because of the complexity, they tend to be stand-alone units, not a probe connected to a DMM."

See my remarks about this below in the Discussion section. It may be, by a "fluke" that our readings ended up being nominally accurate notwithstanding.

The Fluke clamp-on amp meter we were using registered "AC", even though we were measuring DC. As we toggled among the settings, that one was the only one that was in the region that seemed right. The other readings were like 0.11. So yes, the display said it was AC, but it was among the selection for amp readings that included AC and DC. Given that the meter seemed to be of high caliber, and assuming that it worked, I also assumed that its digital readout in the range we were expecting would be accurate, so we went with it. The previous amp meter we borrowed didn't even register. It said "OL" (overload). As it was, it had taken us 2.5 hours to locate the clamp-on amp meter we had, and we had pretty much run out of options.

Premature Shut-Down Because...

I had intended to run the test long enough to see a change in temperature of 20 °C. But as I saw the water in the reservoir going down, and pictured that water (~half a liter?) now sitting in the bottom of the catalytic converter, I pictured the catalytic substrate getting drowned, so I opted to shut off the experiment at midnight when the temperature was 35.5 °C, just 2.5 °C away from the target. I had assumed that most all of the HHO gas would be going to H2O. I was wrong about that. Only about 21% ended up as water in the catalytic converter.

Water in the Catalytic Converter

When we emptied the catalytic converter the next day (March 27), we were careful to be able to catch any water. Frank had drained the tub of its water, so we used that as our work space in case we accidentally spilled while dumping the remaining water into a measuring cup.  It turned out that only 4 ml poured into the measuring cup, as measured by taking the water up into a hypodermic needle.

We weighed the ceramic insulation at the top of the catalytic converter, since it has absorbed some of the water as we tipped the catalytic converter to get the bottom cap off. They weighed 3.45 oz. Frank put them in an oven to dry them out, after which, they weighed 1.25 oz. So the weight of water there was about 2.20 oz, which equates to about 60 ml of water. (A liter weighs approx. 2.204 pounds, or 1000.028 grams.)

We noted that there were a few drips from the emptied catalytic converter as it sat there. We estimated that it was about 1 ml of water.

So the total amount of water we measured in the catalytic converter after the experiment was:

Total = 4 ml + 60 ml + 1 ml = 65 ml

We don't know how much water may have escaped via vapor or steam. With the water in the reservoir being at between 105 and 110 °F, there would be a significant amount of vapor that could pass through. And some of the water at the hot spots of the catalytic converter could be turned to steam. HHO in, catalytic process to H2O, steam out. I wish we had thought to put a bag on the outgoing tube to collect whatever gas was emitted from the catalytic converter.

Water Used by HHO Generator

Seeing that the rate of water consumption in the reservoir was significant, at 7:18 pm, we drew a horizontal line on the reservoir and marked the time: 7:18. We did the same at 11:40 pm. The distance between those lines was around 1/2 inch on the 3 quart container. When the test was over, we took some of the KOH solution we had emptied from earlier when we had the hoses hooked up wrong, put it in a measuring cup, and emptied it into the reservoir to bring it up to the 7:18 line. The amount of water was about 200 ml, and that was the last half of the experiment. See the calculations section for the estimate of total water used by the HHO generator. 



Live Recording: Set-up - 3:36:25


Live Recording: First Half of Test - 5:49:24


Live Recording: Last Half of Test - 2:50:45


1 of 8: Setting Up 


2 of 8: Adding Electrolyte 


3 of 8: First HHO Flow-Rate Test 


4 of 8: Explanation to Kids 


5 of 8: First Data Reading 


6 of 8: Data Reading Protocol  

(random code computerspeak: ah = amp-hour; K = kilo conversion; L = liters; a = ambient; P = power; v = volts; x = times; sU = energy from the universe, which is comprised of You = God; is U)


7 of 8: Evening Observations 

(random code computerspeak: quantum free nrg, rolling out through Y-connections in the HHO tubing) 


8 OF 8: This Week in Free Energy report - excerpt of just the H-Cat - full length

(random code computerspeak: mr Junior = we're just starting out, in the vicinity of 92% efficiency, but the Upcoming is ever so blow away)



Key Time Points

  • March 26, 2014
    • 10:59 am: tub filled with water
    • 3:39 pm: turned on HHO generator to get it warming up
    • 4:44 pm: did gas flow rate test at 10.73 V, 11.55 amps
    • 4:48 pm: started calorimetry test, hooking the HHO gas output to the catalytic converter
    • 5:36 pm: 1) saw that the "+" lead was disconnected; reconnected it and put a clamp on it; 2) started stirring water before taking thermometer readings
    • 8:43 pm: Added packing peanuts on top of the water in the water bath
    • 9:11 pm: "-" connection was smoking, so I turned off the machine for a couple of minutes, put a clamp on it
  • March 27, 2014
    • 12:00 am: stopped the calorimetry test; turned the power supply off; commenced cool-down rate documentation
    • 7:19 am removed packing peanuts from top of water
    • 11:34 am last cool-down temperature reading taken


Here are some graphic representations of the data.

The rise in water temperature during the test, followed by the cool-off period, both seem to be fairly linear. The reading accuracy of the water thermometer was not sufficient to pick up correlations between ambient temperature, or between increases/decreases in input Watts to the HHO generator. There doesn't seem to be any noticeable effect of adding then removing the packing peanuts.

The ambient temperature is shown in pink. It was converted to Celsius, after adding 10F to it in the above chart [that's about how far off Frank's outdoor thermometer is]. The second to last temperature, recorded by Frank, while I was away at a NEST conference call, is probably a toss-out, miss-read. Seeing the gradual drop in ambient prior the start of the test shows why there was a downward temperature pressure on the water before starting.

Note that there is an up and down correlation between the Watt input (pink) minus 100 (to place it in vicinity on the chart) and the resulting water temperature (blue). For the most part, where there is a drop in Watts, there is a slowing of the increase in water temperature. When there is a rapid increase in Watts, there is likewise a larger increase in water temperature.

I wasn't expecting the extent of correlation that we see here, but I'm not surprised. This chart was calculated by the spreadsheet by taking the reading from the Killawatt meter plugged into the mains, from which the power supply was running and dividing it by three (pink dots); and multiplying the volts times amps (= watts, shown in blue) going from the power supply to the HHO generator.

(My reason for dividing the Killawatt meter readings by 3 was due to my insufficiency in dealing with XLS information. To get it so both lines don't look flat, one at the top of the chart and one at the bottom, I had to divide the Killawatt by 3 so you could see the relationship in the movement and shape of the curves using the same scale. So bear in mind the 3x factor.)

This same relationship is shown in the above chart comparing the temperature of the HHO generator (pink dots) with the Watts going into it. There seems to be a delayed effect in cooling following the drop in wattage near the beginning. The last temperature reading on the right makes sense from a vantage point of the power being turned off, but the measurement just before it, while the power was still going to the HHO generator seems errant.

The increase in amperage in each set came because of a better connection to the terminal, first by the "+" lead, then by the "-" lead. Otherwise, each set is nominally horizontal. A similar relationship is shown with Volts in the following chart, except with a slight drop over time (note that the scale below is for very small increments of 0.05V).


Input Watts Versus Output Calories

For enumeration of the calculations used to measure the input Watts in joules versus the output calories in joules, see Sterling's Calorimetry Protocol (March 22).

Since the amps fell into essentially three sets, with two interruptions of the power level between the three sets, I'm going to calculate the input:output joules for those three time slots. It seems to me that the region around when the terminal connections were disrupted is too incomplete to do proper calculation based on the area under the curve. There are not enough data points, so it is best to discard these portions of the data and only go with the more solid three sets.

On the third set, since I didn't take a water temperature reading at 9:13 pm, I'm going to start with the measurement taken at 9:31 pm, but take the average amps and volts from 9:13 to 11:55 pm (both the amps and the volts were quite steady during this set due to both the "+" and "-" terminals being secured tightly in place).

Set A:
Data: From 4:48 to 5:12 pm (24 minutes), with the average amps being 11.59 and volts being 10.68, and with a temperature rise of 0.3 degrees from 18.0 to 18.3 °C.

11.59 amps x 10.68 volts x 24 minutes = 2970 W-minutes
2970 W-minutes x h/60m x kW / 1000 W = 0.0495 kW-h
0.0495 kW-h x 3600000 joules/kW-h = 178000 joules input

0.3 °C x 46 L x 1000 ml/L = 13800 calories
13800 cal. x 4.18 joules/cal = 57700 joules output

Output:Input efficiency: 32%

Note: The duration of this test was very short, causing the temperature difference to be well within the range of error from difficulty of measuring more accurate than 0.25 C. So the results from this calculation are not reliable at all. You'll see in the video of the livestream recording of this region that In the first hour, I calculated an efficiency of 93%, and in the second hour, I calculated an efficiency just over 100%, but that was based on an eyeballing of the amp data and estimating an average in my head. I bring this to recollection to point out that the low reading of the first 24 minutes: 32% is compensated by a high reading if you go with the first hour. A longer measurement period is needed for a more reasonable conclusion. Set B is a little longer, but not much -- nearly two hours. Set C is the longest, covering close to three hours and is the best sampling for a reliable conclusion.

Set B: 
Data: From 5:38 to 7:31 pm (113 minutes), with the average amps being 12.64 and volts being 10.65, and with a temperature rise of 3.9 °C from 20.3 to 24.2 °C.

12.64 amps x 10.65 volts x 113 minutes = 15200 W-minutes
15200 W-minutes x h/60m x kW / 1000 W = 0.254 kW-h
0.254 kW-h x 3600000 joules/kW-h = 913000 joules input

3.9 °C x 46 L x 1000 ml/L = 179000 calories
179000 cal. x 4.18 joules/cal = 748000 joules output

Output:Input efficiency: 82%

Set C: 
Data: From 9:31 pm to 12:00 am (149 minutes), with the average amps being 14.48 and volts being 10.68, and with a temperature rise of 5.5 °C from 30.0 to 35.5 °C.

14.48 amps x 10.68 volts x 149 minutes = 23000 W-minutes
23000 W-minutes x h/60m x kW / 1000 W = 0.384 kW-h
0.384 kW-h x 3600000 joules/kW-h = 1380000 joules input

5.5 °C x 46 L x 1000 ml/L = 253000 calories
253000 cal. x 4.18 joules/cal = 1060000 joules output

Output:Input efficiency: 77%

Water Usage

The usage of water by the HHO generator from 7:18 to 11:40 (262 minutes) was 200 ml. The duration of the test prior to that was 4:48 to 7:18, or 150 minutes, so the estimated usage of water during that time, based on the amount used in the last portion, would be 200 ml/262 minutes x 150 minutes = 115 ml. So the estimated total water usage by the HHO generator comes to 315 ml.

The ratio of water used by the HHO generator (315 ml) versus the amount of water that ended up in the catalytic converter (65 ml) is 21 percent.

Contribution of the Steel Tub

According to
Cv(water)= 1 cal/g.K
Cv(steel)= 0.1 cal/g.K

In other words, it has 1/10 the calorimetric rating of water.

The shipping weight of the tub we were using is listed as 4.25 lbs = 1928 grams. To get liters of water equivalent, we'll use the following information: 1) A liter weighs approx. 2.204 pounds, or 1000.028 grams. 2) The caloric value of steel is 1/10 that of water. So 1926 grams x L/1000 g x 0.1 cal/g.K = 0.02 L. 

In other words, it's essentially negligible, amounting to around 20 ml of water.

Contribution of the Brick

On the other hand, I imagine the contribution of the half brick, weighing 1 lb, 14 oz was more substantial, but since I've not been able to find how to give a rough water equivalent, we'll not factor it in. If I were to guess, I'd imaging it be somewhere less than half a liter of water. 

Factoring in the brick and the Steel Tub would give us slightly better numbers for efficiency.


We first published this report in draft form on March 29. In process of reviewing the lab notes, consulting with experts, and fact checking, there have been a few things that have been modified from the original draft, including:

  • In our initial report, we stated that on the day of the test we measured a 93% output/input efficiency. Upon closer analysis and scrutiny, we have pared that down to 77% as being the most accurate number for this systems' efficiency.
  • Problem: we did not calibrate all of the thermometers we were using, but assumed they were all accurate. Some of them were not.
    Originally, we stated that target water temperature was 55 ΊF. However, in doing the conversion from Celsius to Fahrenheit, and in looking at the readout from the infrared thermometer in Fahrenheit, I realized that the starting water temperature was actually probably closer to 65 ΊF. The Celsius thermometer we used in the water, which read 18.5 °C, agreed with the infrared thermometer. The thermometer I was using when filling the water basin was difficult to read and had very small increments for large temperature differences. It did read 55 ΊF (+/- 1 ΊF), but apparently it was off by 10 °F. 
    At around 12:18 am on March 27, I made a memo: "what I've been calling 49 and 53 may actually be 59 and 63 etc... according to the IR thermometer. So apparently the thermometer we were using to record "ambient" temperature was also off by 10 ΊF.
    Fortunately, the glass thermometer we were using to read the water temperature is most likely very accurate, judging from its design; while the "ambient temperature" thermometer we were using could have very well been inaccurate. The water was nominally in the region of ambient, and the bath was well insulated. The water temperature was the key measurement. All other temperature readings were auxiliary.
    According to the water temperature readings, the starting water temperature was higher than ambient, because it dropped from 18.5 °C at 12:22 pm to 18.0 °C at 4:48 pm when we started the test.
  • Having overlooked a couple of important data points, I originally stated that the water temperature held constant at 18.5 °C from 10:59 until 4:48 when we started the test. The above sentence is what is in my notes. We measured 55 ΊF at 10:59, but didn't get the Celsius thermometer in the water until about 12:22 pm. And the first three readings after starting the test showed 18.0 °C, at 4:48, 4:53, and 5:02 pm. To illustrate the difficulty of reading the thermometer, my notes record that at 4:44 pm, the temperature was observed to be 18.3 °C.
  • Regarding Water Usage: Before I knew what the amount of water in the ceramic insulation was, I had estimated that the total usage was maybe 20 ml, and for some reason I was thinking we had used a total of 400 ml of water in the reservoir, so at one point, I stated that the amount of H20 in the catalytic converter was just 5%. That is incorrect. The actual figure is 21%.
  • Initially, I went with the Wikipedia number for electrolyzer efficiency: 70%. But since Steve D. says his electrolyzer is 75% efficient, then I switched to that number, which changes the maximum conventional efficiency number to 64% (multiplying 75% by the 85% fuel cell efficiency number).


The calculations from the third set of data, showing 77% efficiency, is probably the most accurate due to: 1) having the least variation in volts and amps due to the best connection of the wires to the terminal, 2) longest run time, for the widest variation in temperature and hence reading accuracy.

We didn't get the 2-10x more energy out than what went in that we were hoping for, but we did get quite close to 100% efficiency, which is still very interesting and noteworthy -- especially considering that of the 315 ml of water that was consumed from the HHO reservoir, only ~65 ml was recovered in the bottom of the catalytic converter -- so much of the gas vented out. 

I think we also showed that the probable LENR reaction is probably more prone to take place at higher temperatures than what the below-room-temperature water bath allowed -- given the high temperatures being reported by others, without a water bath. For example, Leon Ward reported achieving 1700 ΊF for at least two hours with an H-Cat.

Think about it, what temperatures do the other LENR researchers have to pre-heat their system to before any LENR takes place? [According to Hank Mills, it's under 100 °C for the Ni-H scenarios.] It's a lot higher than the 65 °F (18.5 °C) we started at. (We were outside in a greenhouse on an overcast day, so our "ambient" was lower than typical "room" temperature, which would be much warmer.)

The other guys experimenting with the H-Cat design (e.g. Neal Ward) are getting upward of 1700 °F (927 °C), but they have not been trying to quantify the input:output ratio.

At a minimum, our results show that the H-Cat can be used as a very efficient DC-to-water-heat conversion. The DC-source we used (old ham radio power supply) was very inefficient, probably producing more heat than the catalyzer.

The temperature graph seems to show that the adding and removal of the packing peanuts did not significantly effect either the heat-up curve or the cool-down curve. The insulation prior to that was apparently sufficient or equally ineffective.

The amperage set changes were due to differing levels of contact of the negative electrode to the power supply. When the connection was fastened securely at the end, the amperage went up into the 14.5 amps range. Prior to that, at 9:11 pm, we briefly turned off the power supply because the connection was smoking. The reading prior to that was 10.7 amps.


Regarding the Fluke meter showing "AC" even though we were supposed to be measuring DC, mentioned above, and the comments saying that this nullifies the results because we were probably just reading some AC artifact through the power supply connected to the mains.

Notwithstanding, I can't help but think that these numbers were close, if not completely accurate, perhaps by a "fluke" (pun intended). I have several reasons for this.

1) See my comments above about measuring the HHO gas flow rate and comparing that to what Steve D's algorithm would predict the flow rate to be, which is based on Michael Faraday's numbers. Our measurement at 4:44 pm with the volt meter showing 10.73; and the clamp-on amp meter showing 11.55 amps, yielded a 0.87 L/minute HHO gas flow rate. Steve's algorithm predicts a flow rate of between 0.7 and 0.75 L/m at that power input. So our numbers are about 15 to 20% higher, but within a range that is reasonable and believable. This shows that the "calibration" factor of the readout is practical.

2) Also, the lowest amp reading: 10.7, occurred just before the connection started smoking because it had gotten too loose. Then, after I put a clamp on the connection, the reading went up to 14.28 amps. That is believable. 

3) Also, at 5:36, seven minutes after our previous reading, we noticed that the + lead had fallen off, so we turned off the machine, hooked it back up with a clamp, and turned it on again. The readings before that were consistent in the region of 11.6 amps. The readings after that were consistent in the region of 12.7 amps. It would make sense that the current would go up with a better connection. The reading just prior to this happening was the lowest of that set, at 11.4 amps. And the reading following this was the lowest for the following set, at 12.14 amps, which would point to the HHO generator warming back up after being shut down for a couple of minutes. This is consistent with what Steve had told me to expect prior to this happening.

This said, I can see that our amp data can be separated into three sets, as illustrated well by the charts shown above.

  1. Beginning, from 4:44 pm to 5:29 pm, Readings: 11.55, 11.56, 11.62, 11.60, 11.63, 11.40 amps
  2. After "+" lead disconnected and was reattached with a clamp, starting at 5:38 to 8:24. Readings: 12.14, 12.72, 12.92, 12.88, 12.62, 10.7 amps (the latter was before the "-" lead started smoking and was then clamped)
  3. After the "-" lead was clamped, from 9:13 pm to 12:00 am, when the test concluded. Readings: 14.28, 14.48, 14.48, 14.61, 14.55 amps

This brings up a fourth reason for believing these numbers could be legitimate.

4) In the three sets of numbers, you see that lowest set was up front, in the region of 11.6 amps. This would have been when both the "+" and "-" leads were not as secure as they were at the end, for the third set.

5) The tight correlation of curves in the Killawatts/3 versus Watts from Power Supply chart above shows that the volts x amps (=Watts) amplitude for the power supply was appropriately proportionate to the amount of power being pulled from the mains. Except for one data point at around 6:44 pm, there is a 100% correlation in direction of movement and magnitude of movement.

6) Ditto for the HHO-Gen Temp versus Watts in chart, though it isn't as tight as the correlation in the Killawatts/3 versus Watts from Power Supply chart.

So, in conclusion, it seems to me that the amp readings we got can be considered legitimate. The calibration of the readout was in a believable range, and the reading value was increased as the connections to the "+" and "-" terminals of the power supply were tightened, and the magnitude increased to a believable extent given fluctuations in HHO-generator warm-up. 

Call it a fluke. I'm going with it.

Also, in further support of something like this being reasonable, let me remind you that my dad, David W. Allan, is an atomic clock physicist. One of the things I've seen him do many times over the years is to get data from places where data is supposed to be impossible. It's his gift. I would argue that GPS would not be what it is today without some of the contributions my dad has made to that field. We all know how practical and valuable GPS has become in the modern world. Google "Allan Variance".

It's the same principle as not just following whatever your Doctor says but becoming responsible for your own health and using your Doctor as a reliable source but not infallible. Sometimes Doctors are wrong. Likewise, sometimes you might have to do flukey things, even with a Fluke meter, as good and reliable as they are, by reputation. It's part of thinking outside the box. Be open.

Comparison to Other Electric Heat

In any system, unless you're harnessing the wheelwork of nature in the process, the output energy is going to be less than the input energy, due to losses.

This equation changes if you are harnessing the wheelwork of nature, in which case there is another "income" represented from the environment, showing up in the system, so that even after the "loss" is factored, there is MORE energy showing up on the "output" side than the "energy input" that was required.

Graphic by Jim Rodney, NEST

The notion of losses in typical systems is well illustrated by this next graphic showing power distribution through the grid.

The question I'm asking is what this ratio looks like for conventional heater systems. I was actually quite surprised by what I found.

According to 

Conversion process

Energy efficiency
Appliance Efficiency
Electric shower 90–95% (overall it would be more efficient to use a heat pump, requiring less electric energy)
Electric heaters ~100% (essentially all energy is converted into heat)

I presume that last number is a reference to electric resistive heaters, where the idea is to dump electrical power into a resistive load in the form of heat.

I should note that the "Electric Shower" number is very interesting because the Steorn product that is poised to be released any time now into the market, is said to use 5 times less electricity than any presently-available conventional electric water heater. This illustrates the principle of harvesting energy from the environment. That's the only way that could happen.

But now, here's the kicker. The following is additional information in that same Wikipedia chart:

Conversion process

Energy efficiency
Electrolysis of water 50–70% (80–94% theoretical maximum)

In the last set of our data, we calculated that we were getting a 77% efficiency, which is approaching the theoretical maximum of electrolysis efficiency. And that doesn't even take into consideration the conventional input:output efficiency of the catalysis of gas back to liquid -- a chemical process that is bound to be inefficient.

Steve D told me that his HHO electrolysis cell is 75% efficient, so we'll go with that number, rather than the 70% max given by Wikipedia above. We got above 75% in the entire system, including the catalytic converter.

What are the numbers for the catalytic portion of this system? From the same Wikipedia chart:

Conversion process

Energy efficiency
Electricity generation
Fuel cell up to 85%
World Electricity generation 2008 Gross output 39%, Net output 33%.

The fuel cell figure alone, says that the highest efficiency we might expect from this catalytic process, which is very similar to a fuel cell, is 85%.

Other References


There is no doubt a catalytic reaction producing heat as HHO goes back to H2O in the catalytic converter. The question we are after is whether or not there might be something else going on as well, whether it be LENR or some other reaction -- probably nuclear, and safe. I think our results definitely point to the latter, even at this low temperature.

We are not computing the power to the power supply as being part of the system since this test isn't about factoring that component, which can vary so much from one manufacturer/model to another. We are limiting it to the DC power going to the HHO unit as representing the "input" power, and the heat being generated by the catalytic converter as the "output" power.

Comparing apples to apples, according to Wikipedia, the highest efficiency in fuel cells is 85%. And according to Steve D, his electrolyzers is 75% efficient. Multiplying these (85% x 75%) to get the combined effect, which is what we see in this system, comes to 64% as the maximum efficiency we can expect using traditional numbers for systems such as this. So our measured in the region of 77% efficiency is encouraging.

On top of that, when you consider that ~21% of the water that became HHO gas ended up as liquid H2O in the catalytic converter chamber, minus whatever vented off as water vapor or steam; I think it's definitely safe to say that we are getting help from some anomalous source, such as LENR.

Furthermore, bear in mind that according to the calibration of the power readings according to Steve D's algorithm of L/minute HHO gas flow rate, mentioned above, our numbers for the volts x amps were about 15-20% high. Based on our amps and volts, his algorithm predicted an output rate of 700 to 750 milliliters per minute. We measured 0.87 L/m, which is 15-20% higher than what he predicted, which means that the ACTUAL amps were higher in order to achieve that higher rate of flow. So technically, by this calibration adjustment factor, we could say that the ACTUAL efficiency was 15-20% HIGHER than 77%. But because we didn't calibrate it across several power settings, and our most reliable data came from the third set, which was substantially higher than the first set, in the region where we did the L/m flow test (that ended up serving for calibration), we can't really use this adjustment factor in any reliable way, other than to say it argues in favor of an even higher efficiency, and thus in favor of an anomalous effect, not against.

Also, given our observation, which Steve sees routinely, that the water reservoir was substantially warmer than the generator; along with my being told a decade ago that Brown's gas has tritium in it, I can't help but think that this could be contributing both to the warmer reservoir as well as the heating effect in the catalytic converter, that is far above what conventional science would predict.

In my opinion, these factors all point to some kind of safe nuclear reaction taking place in the catalytic converter. 

There is definitely an anomalous effect taking place here, even at below room temperature.

While 77% efficiency might not compete with existing electric heaters, in this context of an electrolyzer and catalyzer, it points to an effect that should be investigated, characterized, optimized, and implemented for energy generation worldwide.

Graphic by Jim Rodney, NESTThis illustrates a very important principle. Something doesn't have to be technically "overunity" -- meaning over 100% efficiency -- in order to be manifesting the harnessing of the wheelwork of nature: plugging into a principle of nature to assist the output/input ratio. And if it is showing up here, what will happen when we characterize and optimize it? (Graphic by Jim Rodney, NEST)

As far as products go, we've shown that heating a water bath using this method isn't the best use of this phenomenon, at least not with the ingredients we used. There are sure to be better ways to put this effect to practical use.

Someday, we expect to see paradigms like this making it to market. This H-Cat could get there fairly soon.

Next Time: Suggestions

We should have used a rubber band to fasten an empty plastic bag to the end of the hose coming from the catalytic converter, to capture and nominally quantize the exiting gasses and water vapor. If it filled, we could have sealed it and replaced it with another bag. Given that only 21% of the water that was turned to HHO ended up as H2O in the bottom of the catalytic converter, I really wish we had taken this step to try and quantize the output gas. I'm guessing that if there was steam or water vapor that escaped that some of it would have gone back to liquid water in the bag. Mike Waters had suggested that we run the exiting gas into a soapy solution to ten try to ignite the bubbles. If it was predominantly HHO, it would have ignited easily.

I should have taken a gas flow rate measurement at the end, disconnecting the HHO-gas tubing from the catalytic converter (to stop the reaction) and running it into our 2-liter bottle to time how long it took to fill it at that volt x amp setting. Our Volts x Amps reading was quite a bit higher at the end than it was at the beginning. It would have been nice to know what it was. Then we would be able to get a calibration from Steve's algorithm in that last of three sets of data.

I should have measured the temperature of the exiting gas.

We should have calibrated all instruments, including thermometers, volt meter, amp meter.

I wish we would have recorded the reservoir temperature throughout the test as well. That was a very interesting reading, being consistently several degrees higher than the warmest spot on the HHO generator. It didn't fluctuate much, but it is interesting and its significance may become more clear later.

I wish I had videotaped us emptying the water from the catalytic converter. That was quite stunning to be expecting about half a liter of water but only getting about 10 ml. 

I also wish I had taken more photos of the whole set-up -- system photos. I don't have even one photo that shows everything in one photo in its operational form.

While we don't have a data logger, someone in the comments suggested we could have a video camera trained on each of the meters for the duration of the test, then manually enter the numbers. If the fidelity rate of the camera were decreased, it could videotape longer before having to have a sim card changed, etc. 

Design Proposal: Basic Wind Tunnel

Mike Waters posed an interesting idea during our NEST meeting March 27. He thinks the LENR is much more likely to take place at higher temperatures, so what we need is a way to maintain a higher temperature and measure the input v. output.

He proposed basically a simple wind tunnel with the following components. 

  • Use a speed-controllable fan so you can adjust the wind speed.
  • Use a wind-speed meter (cheaply and readily available from wind turbine suppliers). (Several would be preferable, to then average or otherwise integrate.)
  • Insulate the walls of the tunnel.
  • Use an infrared laser to check the temperature of the H-Cat, then adjust the wind speed to maintain that temperature.
  • Recycle the outgoing water and gas back into the system, since it is likely to contain things like deuterium, tritium, and other things that would be desirable to accumulate rather than discard.
  • (Given all the variables likely to change, this system is going to require a data-logger to record data in real time, so the data can be tabulated later via algorithm.)

My dad doesn't think the math for this is going to be straightforward nor very accurate, since you'll lack uniformity of air flow. Fluid dynamics are not easy to tabulate.

Design Proposal: Heat Exchanger:

I'm thinking perhaps a better method than the wind tunnel would be to have ThunderBoltPerfromance embed a heat exchanger in a catalytic converter matrix embedded with nano-palladium and nano-nickel; and then insulate around that very well, for high-temperature testing, with high-temperature thermal probes, and run a high-temperature fluid through it from an insulated bath, whose temperature is measured to rise over time enough to get a good delta T. So it would be the same set-up as the water bath we just did, except that we would regulate the fluid flow rate to maintain a given temperature in a higher range. 

What kinds of fluids would be good for this? What would the best flow methodology be on a limited budget? I would think that targeting a temperature range, then setting a given flow rate, would be best for ease of measurement. That way' you're just measuring volts and amps to the HHO generator, and the change in temperature over time to calculate watt-hours and calories of heat.

Hank Mills says that the minimum temperature required in the Nickel based LENR reactions by E-Cat, Defkalion, and Brillouin, is below 100 ΊC. That's not practical using water because it's in the region of phase change, but it could be easily achieved and maintained with another fluid with a higher boiling point.

In Other News

Speaking of highly efficient electric water heaters, Steorn's claim is that their product, due to roll out any time now from two of the largest water heater companies in the world, will cost around as much as competitor electric water heaters, but will consume 1/5 as much energy. (Story) Per the principles discussed above, that points to having to harvest the wheelwork of nature in the process, in order to achieve that. 

Bear in mind that when a  licensee announces their new product, they're not going to say "brought to you by Steorn". It's getting close to the time that this product should be released, so we should be on the lookout for it. Which of you will spot it first?

See also: Weird Resonance in H-Cat Heat Test 

And H-Cat News


Click here to donate.

We appreciate those of you who have been sending us donations. While we do have a couple of major finance deals percolating, those take time, and meanwhile we need to stay afloat. Some day soon, a lot of us (including most of you reading this) are going to be growing very wealthy as we help the planet be a better place by providing products and services they need to implement exotic free energy solutions. Until then, thanks for chipping in, to help us bridge the gap prior to arriving at that point. Justin Church likewise needs support.

Donations to PES will be used both for this project as well as for maintenance of survival. We've got some heavy expenses we're under right now, so we appreciate those of you who are chipping in.

Click here for tally of amount raised so far.

# # #

What You Can Do

  1. See Suggestions for How to Get Involved with the Roll-out of Exotic Free Energy 
  2. Pass this on to your friends and favorite news sources.
  3. Donate to PES Network to help us keep this news and directory and networking service going.
  4. Subscribe to our newsletter to stay abreast of the latest, greatest developments in the free energy sector.
  5. Let professionals in the renewable energy sector know about the promise of this technology. 

See also

Resources at

Page posted by Sterling D. Allan Jan. 29, 2011
Last updated April 26, 2014




"It is harder to crack a prejudice than an atom." // "I'd rather be an optimist and a fool than a pessimist and right." -- Albert Einstein

ADVISORY: With any technology, you take a high risk to invest significant time or money unless (1) independent testing has thoroughly corroborated the technology, (2) the group involved has intellectual rights to the technology, and (3) the group has the ability to make a success of the endeavor.
All truth passes through three stages:
   First, it is ridiculed;
   Second, it is violently opposed; and
   Third, it is accepted as self-evident.

-- Arthur Schopenhauer (1788-1860)

    "When you're one step ahead
of the crowd you're a genius.
When you're two steps ahead,
you're a crackpot."

-- Rabbi Shlomo Riskin, (Feb. 1998)

Submit • Privacy • About • Contact

PESWiki Departments:
Latest • News •XMLFeed • Directory • Congress • Top 5 • Open Sourcing
Copyright © 2002-2015, PES Network Inc.