Other Thunderstorm Generator tests: who ran each one
Several tests of the Thunderstorm Generator have been published. This page lists each one with who ran it, who paid, what was measured, and what the result was. The university thesis is shown first. The seller-side tests come second and are labeled as reported by the inventor’s side.
Measured
Measured: the university thesis
These three tests are from one master’s (M.S.) thesis at Embry-Riddle Aeronautical University (2026). For the road test, the main vortex unit, piping and bubbler were supplied by Plasmoid Power, the seller.
| Test | Who ran it | Who paid | What was measured | Result |
|---|---|---|---|---|
| Idle exhaust test, small generator, 10 runs each way | Walker Hall, M.S. student, with faculty advisers named in the thesis | Not stated in the thesis text we read | Exhaust gases from a Troy-Bilt 5550 generator at idle inside a trailer, read with a five-gas analyzer | Carbon monoxide (CO) down by an average of 34%. Carbon dioxide (CO₂) and oxygen: no significant change. |
| Fuel test, small generator under load (about 2,760 W) | Same | Not stated in the thesis text we read | Gasoline weighed and timed on each run, baseline versus retrofit with the water-vapor (cold-gas) system on | Fuel-use rate 35.27 mL/min down to 31.86 mL/min, 9.67% lower |
| Road test, 1989 Chevrolet Blazer, 10 runs each way | Same | Not stated in the thesis text we read | Fuel tank weighed before and after a 32.2 km (about 20 mile) loop | About 8.7% better kilometers per liter. Only 4 of 8 cylinders were covered, and retrofit runs had high winds. |
Inventor claim
Reported by the inventor’s side: four tests
These four tests are published on Thunderstorm.tech, the seller’s site. We have not seen the full original reports. We list them as reported by the inventor’s side, not as measurements we can vouch for. The site describes them as independent and certified. We could not check who paid for any of them.
| Test | Who ran it | Who paid | What was measured | Result as reported |
|---|---|---|---|---|
| Dartford, UK, June 28, 2024 | Element, a testing company, per Thunderstorm.tech | Not stated on the page | Exhaust gases from a Thunderstorm Generator-fitted engine, with and without the device. Engine type not given. | Carbon monoxide (CO) −97.1%, volatile organic compounds (VOCs) −99.9%, sulfur dioxide −84.8%, nitrogen oxides −6.5%. Oxygen 0.15% to 10.40%. CO₂ is not in the table. |
| Clearwater, Florida, June 24, 2024 | Not named on the page | Not stated on the page | Exhaust gases and fuel run time, base-load generator at 60 Hz under a 2,400 W load | CO₂ −99.56%, CO −98.00%, nitrogen oxides −95.14%, hydrocarbons −77.47%, fuel run time +19.45%. Units are unclear in the page’s table. |
| Star Cement, United Arab Emirates (UAE), December 5–6, 2024 | Described as run in partnership with Strike Foundation (UK) | Not stated on the page | Ambient air, plain exhaust and treated exhaust on a Honda 5 kW gasoline engine | CO₂ −83.33%, hydrocarbons −77.30%, nitrogen oxides −79.66%. CO unchanged at 0.10%. Oxygen 9.12% to 20.43%. |
| Plasmoid Power’s own test program (Test 13) | Plasmoid Power, signed by its chief executive; the page calls it self-certified | Not stated on the page | Five-gas analyzer on a 5,500 W generator under a 2,200 W load, five runs over four hours | CO₂, CO and nitrogen oxides “almost eliminated,” oxygen near air level. Hydrocarbons rose. The first run was left out of the averages. |
Plain definition
How the results compare
The numbers differ a lot from test to test. CO fell by 34% in the thesis, by 97–98% in two seller-side tests, and not at all in the UAE test. CO₂ and oxygen did not change in the thesis, but moved sharply in the seller-side tests.
Different engines, loads, analyzers and test methods can explain differences. They also mean no single number tells the whole story. A gas analyzer reports concentration, which is how much of a gas is in each sample. It does not by itself show how much gas the engine made.
Evidence timeline
Tests and sources this site cites, oldest first. Open an item for the details and the source.
Wide audience: the Joe Rogan ExperienceInventor’s side
Randall Carlson introduces Bendall’s work on episode #1897 of the Joe Rogan Experience, as Strike Foundation tells it (it dates the episode Nov 11).
Strike Foundation — home page (inventor’s side)Joe Rogan, Carlson and Bendall
Clearwater, Florida generator testSeller claim
Reported: carbon monoxide (CO) −98.00%, carbon dioxide (CO₂) −99.56%, fuel run time +19.45%. Tester not named.
Thunderstorm.tech — Test Results (seller’s site)All tests side by side
Dartford, UK test (Element, per the seller)Seller claim
Reported: carbon monoxide (CO) −97.1%, oxygen 0.15% to 10.40%. Full report not seen.
Thunderstorm.tech — Test Results (seller’s site)All tests side by side
Star Cement, United Arab Emirates (UAE)Seller claim
Reported: carbon dioxide (CO₂) −83.33%, carbon monoxide (CO) unchanged at 0.10%, oxygen 9.12% to 20.43%.
Thunderstorm.tech — Test Results (seller’s site)All tests side by side
Skeptoid critiqueOutside commentary
Brian Dunning argues the device is an exhaust heat riser plus water injection; small gains are “probably true” if they stay small.
Skeptoid #1019, Brian Dunning, Dec 16, 2025Critics and supporters, side by side
Embry-Riddle master’s thesisMeasured
A 9.67% lower fuel-use rate on a small generator; carbon monoxide (CO) about 34% lower at idle; CO₂ and oxygen not significantly changed; about 8.7% better economy on a one-vehicle road test using hardware the manufacturer supplied.
Hall, W. (2026), Embry-Riddle Aeronautical University master’s thesisWhat the test measured
Plain definition
What we could not verify
We could not open the full Element report, and we do not know who paid for it. The Clearwater page does not name who ran the test. We did not find the original data behind the UAE trial beyond the summary table.
If you hold a full report, send it to us and we will read it and update this page.
Sources
Pages we opened and read for this explainer. The label in parentheses says which side each one speaks for.
Get new explainers by email
About one short email when a new explainer or correction goes up. No ads, no selling your address. Unsubscribe anytime with the link in every email.