"specific, substantial and credible"
A patent is a disclosure and exclusion right
A patent record can be valuable. It can show dates, named inventors, diagrams, claimed mechanisms, cited prior art, and what the applicant said the invention does. For historical research, those records are often worth reading closely.
But a patent is not the same as a peer-reviewed measurement report. It does not normally include an independent laboratory energy balance. It also does not mean a commercial product can deliver the claimed performance under controlled conditions.
Utility is not the same as over-unity proof
U.S. patent examination requires utility that is specific, substantial, and credible. That screens out many impossible or unsupported assertions, but it is still a patent-law standard rather than a complete scientific validation program.
The distinction matters when someone says, "It has a patent, therefore it works." The stronger question is: where is the boundary diagram, calibration record, raw input-output data, uncertainty budget, and independent replication?
Perpetual-motion claims can trigger model scrutiny
USPTO examination rules recognize that working models may be requested in some cases, including perpetual-motion situations. That is not a general rule that every patent is built and tested. It is a safeguard for exceptional claims.
If a claim would overturn the first or second law, the evidence burden does not end at the patent office. It begins there.
How to read a free-energy patent
Start with the claims, not the title. The claims define the legal boundary. Then read the specification for what is actually measured, whether the examples report energy input and output, and whether the cited test conditions could rule out stored energy or hidden inputs.
Look for quantities: volts, amperes, power factor, watts, joules, time, flow rate, pressure, mass, temperature, product composition, and uncertainty. A patent that lacks those numbers can still be historically interesting, but it should not be used as net-energy proof.
FAQ
Does the USPTO build every patented device?
No. Patent examination is mostly document-based. Models may be requested in particular situations, but a patent record is not a universal working prototype certification.
Should free-energy patents be ignored?
No. They can be useful source documents. They should be treated as claims and disclosures, then checked against measured evidence.
Cite this page
Free Energy Research. "Patents Are Not Proof of Net Energy." Updated 2026-07-06. Accessed from https://freeenergyresearch.org/patents.
https://freeenergyresearch.org/patentsPrimary sources
- Utility Guidelines Training Materials United States Patent and Trademark Office
USPTO guidance on the specific, substantial, and credible utility requirement.
- MPEP 608.03, Models, Exhibits, Specimens United States Patent and Trademark Office
Patent examination rule noting when a working model may be required, including perpetual-motion situations.
- First Law of Thermodynamics NASA Glenn Research Center
Plain-language statement of energy conservation and energy accounting for thermodynamic systems.