Earth Rotation Power

Speculative If the 17.3 ± 1.5 µV of direct current measured across a manganese-zinc ferrite shell at Princeton is what its authors say it is — and no independent laboratory has repeated it, which is what section 7 asks for — then there is a way to draw electrical power from Earth’s rotation with no fuel, no moving parts, no weather dependence and no siting constraint beyond pointing the device correctly. That conditional is the whole proposition, and it goes first because everything after it is arithmetic on a number one group has measured once.

Established The reservoir is not the constraint: Earth’s rotational kinetic energy is about 2.14 × 10²⁹ J, and tidal friction already withdraws from it at about 3.7 TW, lengthening the day by roughly 2.5 ms per century. A civilisation drawing power from rotation at the rate the tides already dissipate would roughly double that drift, to about 5 ms per century — a number Earth-rotation services would notice and nobody else would.

Frontier The withdrawal rate is the constraint, and the demonstrated rate is 1.1 × 10⁻¹³ W. The measured 17.3 µV and 25.4 ± 1.5 nA imply a source impedance near 681 ohms, and the most a linear source with that open-circuit voltage and short-circuit current can deliver into a matched load is a quarter of their product. The shell occupies 9.4 × 10⁻⁵ m³, so the demonstrated power density is 1.2 × 10⁻⁹ W per cubic metre. At that density a one-gigawatt station is a solid cube of ferrite about 950 km on a side, and the paper declines to estimate how far the density might improve.

Frontier So the prize actually on the table is a physics result, it goes to whoever runs the replication, and it costs under $10,000. That is the lead author’s figure for reproducing the apparatus; the original build cost about $25,000. What is at stake is whether a textbook impossibility argument — that a conductor sitting still on a rotating Earth can draw no net electromotive force from the part of the geomagnetic field that does not rotate with it — has a real loophole. Speculative Everything beyond that, including every watt, waits on a scaling law nobody has demonstrated and on the ceiling in section 5 that the mechanism’s own governing condition imposes.

Established A conductor moving with velocity v through a magnetic field B feels a force per unit charge of v × B, and that is the basis of every generator ever built. A device bolted to the ground in New Jersey is moving: Earth carries it eastward at 354 m s⁻¹ at Princeton’s latitude, about 465 m s⁻¹ at the equator.

Established Part of the geomagnetic field does not rotate with Earth. The field is a core dynamo product, approximately dipolar, axis tilted about 10 degrees from the spin axis, surface intensity 30,000 to 60,000 nT. Its axially symmetric (m = 0) part is by construction unchanged by a rotation about that axis, so it is stationary in the rotating frame in the only sense induction cares about. Chyba, Hand and Chyba sum the m = 0 dipole, quadrupole and octupole terms to get B = 45 µT at Princeton, dipping 57.5 degrees below the horizontal.

Established And yet the standard answer for two centuries is that no current can be drawn from this, for a reason that is not an accident of engineering. If v × B is curl-free inside the conductor — as it is for uniform v and uniform B in a magnetically inert medium — it is the gradient of a potential. Charge redistributes until an electrostatic field cancels it exactly, the net electromotive force around every closed circuit is zero, and nothing flows however large the device. This is the argument that made the subject disreputable, and it is correct as far as it goes.

Frontier Chyba and Hand’s 2016 claim is that the argument assumes a magnetically inert medium, and that a strongly permeable one breaks it. Inside a material with relative permeability mu-r far above one, the internal B is reshaped by the material; choose the geometry so that the curl of v × B is non-zero within the material and the cancellation is no longer exact.

Established The two conditions this imposes are specific and were both tested. A topological one: the body must be a shell. The authors state the predicted emf “= 0 if a = 0 (as must be so since ∇ × (v × B) = 0 for a solid cylinder), or if mu-r = 1”. And a condition on scale and material: the magnetic Reynolds number R_m = sigma × mu × v × xi must be below 1, so the field diffuses through the material rather than being dragged with it.

Established The resulting device is an M100 manganese-zinc ferrite cylinder 29.9 cm long, outer radius 1.0 cm, inner-to-outer radius ratio 0.61, its long axis orthogonal to both the local velocity and the field. The authors measured mu-r = 9,500 ± 2,850 and sigma = 2.07 ± 0.22 S m⁻¹ on their own sample, giving R_m = 0.088 ± 0.028. MnZn ferrite is the unusual pairing the scheme needs: very high permeability with very low conductivity, which is how a 2 cm object lands two orders of magnitude below the Reynolds ceiling.

Frontier The prediction was −13.7 ± 7.2 µV and the measurement 17.3 ± 1.5 µV. Those agree in the sense that the intervals overlap, and it is a weak test: a prediction carrying 53 per cent uncertainty, driven by a 30 per cent uncertainty on a permeability that enters squared, cannot be confirmed to better than a factor of two by any measurement whatsoever.

Speculative If the device is anything, it is a brake on Earth, not a tap into nothing. The bookkeeping requires a retarding force, and at the demonstrated power it is a fraction of a femtonewton. Handwave Any claim that the energy provenance has been demonstrated, either way, is assertion: this experiment measures a voltage, not a source.

Frontier The competing explanation is not exotic and lives in the same material. MnZn ferrites are respectable thermoelectrics; the authors measured a Seebeck coefficient of −417 µV K⁻¹ on their sample, so a 41 mK end-to-end difference reproduces the entire claimed signal. The paper’s central move is to separate the two by their signatures — thermoelectric voltage tracks the gradient, the claimed effect tracks orientation — and whether that separation holds is what section 3 works through and section 7 would settle. Handwave Nothing here supports the phrase “power source”.

Established The measured record is one positive result, one published null, two published theoretical refutations, one published reply and zero independent replications. Its shape — one group, one apparatus, a microvolt signal, a live alternative explanation — is the situation in which replication rather than argument decides.

Established The proposal began as theory with no experiment. Chyba and Hand published in Physical Review Applied in July 2016, and the abstract is candid: “There is a simple general proof that this is impossible. However, we identify a loophole in that proof and show that voltage can be continuously generated in a low-magnetic-Reynolds-number conductor rotating with Earth, provided magnetically permeable material is used.” No measurement and no power figure — only that the effect “requires testing and further examination to see if it can be scaled”.

Established The first experimental test was a null and it was not close. Veltkamp and Wijngaarden reported in 2018 that “the energy harvested is more than 3 orders of magnitude smaller than predicted”. Frontier That null is routinely cited as settling the matter and does not: the 2018 antenna was a different configuration, and the 2025 paper holds that earlier attempts did not satisfy the topological and low-Reynolds-number conditions. Whether that is a fair reading or a moving target is a judgement to make with both papers open.

Established The theoretical objection is old, specific and published in the same journal. J. Jeener of the Universite Libre de Bruxelles posted a refutation as a preprint and published it as a Comment in 2020, unhedged: “no device of the proposed type can produce continuous electric power, whatever its configuration or size”. His core charge is that the positive prediction follows from “a confusion of frames of reference at a critical step” and vanishes when corrected.

Frontier The reply turns on a feature of the geomagnetic field rather than on algebra. Chyba and Hand hold that Jeener’s conclusion “results from a failure to recognize a distinction between the corotation with Earth of the nonaxially symmetric components of Earth’s magnetic field and the nonrotation with Earth of the axially symmetric components”, and that “because the distinction has considerable empirical support, Jeener’s argument is likely incorrect”. Established The distinction itself is not controversial; whether it licenses the loophole is.

Established The measurement appeared in Physical Review Research on 19 March 2025 after a review of nearly two years: received 9 April 2023, accepted 21 January 2025, corrected 28 May 2025. The authors are Christopher F. Chyba (Princeton), Kevin P. Hand (Jet Propulsion Laboratory) and Thomas H. Chyba (Spectral Sensor Solutions, LLC). The headline numbers: “Regression over the data for emf at the orientations of 0 and 180 degrees yields 17.3 ± 1.5 µV” and the corresponding current “25.4 ± 1.5 nA”. Secondary coverage rounds this to 18 µV; this brief uses the paper’s figure.

Established The control battery is the strongest part of the paper. Four predictions became four tests. Rotating to 90 and 270 degrees should give zero, and did. A further 180 degrees should reverse the sign, and did. An “otherwise-identical solid MnZn ferrite cylinder generates zero voltage at all orientations”, as the topological condition demands. And a high-Reynolds-number MuMetal shell gave “emf = 0.019 ± 0.099 µV at 0 degree orientation” — a clean null at a fiftieth of the claimed signal.

Frontier The “second location” is a site-noise control, not a replication, and that distinction is the most important thing in this section. The authors “reproduced our experiments in a residential building 5.5 km east of our original experimental setting”: same latitude, same field, same apparatus, same operators, same analysis. It rules out a building-specific artefact, cannot rule out an apparatus-specific or method-specific one, and the paper does not claim otherwise.

Established The thermal control is where the argument is won or lost, and the paper does it by signature rather than by subtraction. Both ends were instrumented at a stated 0.01 °C precision; shell temperatures “varied from 18.9 °C to 21.5 °C”; data were “recorded on all three multimeters every 10 s for 5–10 h periods”. Plotting emf against the end-to-end temperature difference, a pure thermoelectric voltage falls on a line of slope equal to the Seebeck coefficient, while the claimed effect appears as a vertical displacement of that line — opposite in sign at 0 and 180 degrees, absent at 90 and 270. That is a real discriminator.

Frontier It is a discriminator with one specific weak point. At −417 µV K⁻¹, a steady 41 mK end-to-end difference — four times the quoted temperature precision, small against the 2.6 °C range the shell saw — reproduces the whole 17.3 µV. The method survives that only if the thermal environment is uncorrelated with orientation. But orientations were reached by physically rotating the apparatus, each run “begun by rotating the shell to its appropriate position and then allowing the system to sit overnight”, resuming “in the morning following a period of at least 8 h”. Rotating an object changes how it couples to a room. The paper describes no randomisation of orientation order and no blinding. Frontier It also reports a residual it does not explain: “a systematic offset upon reversing polarity of 1.20 ± 0.35 µV”, seven per cent of the claimed effect. Reporting it is to the authors’ credit; it is also the size of the thing a replication has to beat.

Established The theoretical objection to the 2025 result is on the record too, and it sharpened between drafts. Iver H. Brevik, Moshe M. Chaichian and Mikhail I. Katsnelson posted a Comment on 29 September 2025 and revised it on 23 March 2026, holding that the derivation omits “the electromagnetic boundary condition on dielectric surfaces in motion” and that “the source of the discrepancy... comes only from the imposition of the boundary conditions”. Frontier Their first version argued from the Maxwell stress tensor that “a displacement of the shell in the mentioned y direction, does not involve any expenditure of external work”; the revision shows that with the boundary conditions properly applied “the total power given to the shell will be zero”, and that the Chyba-Hand local dissipation values “oscillate and become negative in a large region of phi around the shell, what is physically not acceptable”. Established A negative local dissipation is not a matter of taste; it is a diagnostic that something in a derivation is wrong.

Frontier The physicists asked to comment split, and they split on the measurement rather than the theory. Yong Zhu of Griffith University said “there are so many factors that can produce microvolt signals”, naming stray capacitance and eddy currents, and that “ruling out all these possibilities will require more experimental evidence”. Carlo Rovelli said “maybe there is a subtler version of the argument that rules out this possibility; I do not know”. Wijngaarden, who ran the 2018 null, remains “still convinced that the theory... cannot be correct” and notes that “the observed voltages are so small that there are many potential spurious causes available”. Paul Thomas, emeritus at Wisconsin-Eau Claire, called the work “very convincing and remarkable”.

Established The authors are the clearest-eyed people in the story, which is worth recording in a corpus that mostly has to correct proponents. Chyba has said the result “has a whiff of a perpetual motion machine”, that at this scale “it is very easy to make mistakes”, and that of scaling “none of that has been demonstrated, and it might well prove not to be possible”. On what happens next: “some independent group needs to reproduce — or rebut — our results, with a system closely similar to our own”. Established As of the most recent reporting nobody is doing it: Chyba told The Debrief he was “not currently aware of any other group trying to reproduce their results”. The build cost about $25,000; he puts a reproduction at under $10,000, the ferrite cylinder at $400 and the multimeters at about $1,500 each.

Frontier What the paper does not contain is as load-bearing as what it does. No torque measurement, no spin-down measurement, no mechanical back-reaction, and therefore no direct evidence about where the energy comes from. No blinding, no preregistration, no randomisation. One apparatus. One latitude. Speculative Every one of those gaps is cheap to close, which is the most encouraging thing about the subject.

Speculative One prediction the paper implies but does not state is the sharpest test available, and it falls out of two of its own equations. The predicted emf is proportional to v, to a field-dependent coefficient, to shell length and to R_m²; and R_m = sigma × mu-r × mu-0 × v × b is itself linear in v. Taken at face value the emf goes as the cube of the local surface speed, softened by a logarithm, and linearly in the axisymmetric field. Frontier Since surface speed falls as the cosine of latitude while the axisymmetric field strengthens toward the poles, that is a steep, easily testable latitude law — arithmetic on the published equations, not a claim the paper makes. Section 7 turns it into numbers.

Frontier Question one is whether the 17.3 µV exists outside this apparatus, and nothing else is decidable until it is answered. A signal seven per cent above its own unexplained polarity systematic, from one group with one cylinder at one latitude, is a hypothesis about an apparatus before it is a hypothesis about the Earth.

Frontier Question two is whether the impossibility argument has the loophole claimed, and this one is decidable on paper. Jeener says the positive prediction comes from a frame-of-reference error; Chyba and Hand say Jeener misses the corotating and non-corotating decomposition of the field; Brevik, Chaichian and Katsnelson say both miss a boundary condition at a moving dielectric surface and that fixing it zeroes the delivered power. Established Three parties are doing the same calculation and getting different answers, and at most one of them is right.

Established Energy conservation cannot referee this experiment, and saying so precisely is more useful than invoking it. Supplying 1.1 × 10⁻¹³ W at 354 m s⁻¹ requires a retarding force of 3 × 10⁻¹⁶ N — three tenths of a femtonewton on a ferrite cylinder standing in a room at 20 °C. Handwave So any argument of the form “it must be wrong because the energy has to come from somewhere” is untestable on this hardware: the energy comes from somewhere on either account, and the mechanical signature that would separate them is ten orders of magnitude below detection.

Frontier Question three is whether the effect, if real, is the claimed one or a different mechanism with the same orientation signature. Any competitor has to reproduce a sign flip at 180 degrees, a null at 90 and 270, a null for a solid cylinder and a null for a high-Reynolds-number shell. Speculative A thermal mechanism tracking the apparatus’s orientation relative to a wall, a duct or a window would reproduce the first two and not obviously the last two, which is why the co-located dummy in section 7 is the control that matters.

Speculative Question four is whether the mechanism has a hemisphere signature. The axisymmetric field points into the ground in the north and out of it in the south, so for a device in the same orientation relative to local vertical and local meridian the sign of v × B reverses across the equator. A real geomagnetic effect must flip; an artefact has no reason to.

Frontier Question five is the one the proposers care about and has the least evidence: does anything scale? The paper gestures at “decreasing b and a while increasing the values of sigma and/or mu (were materials with such characteristics available or created) or v (for example, in an orbiting satellite)” and at many units “physically... in parallel but connected in series”. Handwave It gives no number and says scaling “will be the subject of subsequent investigations”. Section 5 does the arithmetic the paper declines to do.

Frontier Question six is what a null would establish. A well-instrumented replication seeing nothing at the 1 µV level would not refute the 2016 theory; it would refute the 2025 measurement, which is a different and more useful thing. Established Anomaly Resolution in Physics records that across six high-profile discrepancies at three sigma or more since 2010, the number that turned out to be new physics is zero.

Established The first bottleneck is metrological: the entire claim is a 17.3 µV DC displacement in a system whose own material has a Seebeck coefficient of −417 µV per kelvin. That is a thermal sensitivity twenty-four times the signal per kelvin of gradient. The apparatus must hold, or at least measure, end-to-end temperature differences at the tens-of-millikelvin level, indefinitely, through mechanical reorientation, in a room that varied by 2.6 °C over the campaign.

Established The second is that everything in the measurement chain is also a thermocouple. Every junction of dissimilar metals — electrode to wire, wire to binding post, post to instrument — generates its own thermoelectric voltage under a gradient. The reported “systematic offset upon reversing polarity of 1.20 ± 0.35 µV” is the visible residue of that class, and it sets the floor an independent measurement has to resolve.

Frontier The third is the theory’s own precision. The predicted emf depends on R_m², and R_m is linear in permeability, so the 30 per cent uncertainty on mu-r = 9,500 ± 2,850 enters squared. A theory that can only say “between about 7 and about 21 microvolts” cannot be strongly confirmed by a measurement of 17.3, nor strongly refuted by one of 10.

Established The fourth is the mechanism’s own governing condition, and it is a hard ceiling rather than an inconvenience. The effect is claimed only for R_m below 1; the demonstrated device runs at 0.088. Since the predicted emf goes as R_m² times a logarithm that shrinks as R_m grows, pushing right up against the ceiling buys a factor of about thirteen in voltage and no more — roughly 220 µV from a 30 cm unit. Speculative That is this brief’s arithmetic on the published equations rather than a figure the authors give, and they give none; treat it as an order of magnitude.

Frontier The fifth is that the source impedance does not improve with the trick that improves the voltage. The measured 17.3 µV across 25.4 nA implies about 681 ohms internal, in a material of conductivity 2.07 S m⁻¹ — roughly seven orders of magnitude worse than copper, and not incidental, because low conductivity is what keeps the Reynolds number below unity. The mechanism requires a bad conductor, and a bad conductor is a bad generator winding.

Frontier The sixth is the resulting power density, which decides whether this is an energy subject at all. A quarter of the product of open-circuit voltage and short-circuit current is 1.1 × 10⁻¹³ W from a bounding volume of 9.4 × 10⁻⁵ m³: 1.2 × 10⁻⁹ W per cubic metre. One watt needs about 9 × 10¹² units occupying about 0.85 km³; one gigawatt is a cube of ferrite roughly 950 km on a side. Grant the full thirteen-fold voltage ceiling and a proportionate current gain and the cube is still about 175 km on a side.

Handwave There is no known route from those figures to a power system, and the power case is not thin but absent. Nobody has published a scaling law, an alternative material with a higher permeability-conductivity product below the Reynolds ceiling, or a geometry that beats the shell. Speculative The one variable with real headroom is v, and the paper names the way to raise it: put the device in orbit, where it no longer corotates with the field — at which point the claimed effect is unnecessary, because ordinary motional induction already works there.

Frontier The seventh is sociological and it is the binding one now. The replication costs under $10,000 and takes months, and has not happened. Every other bottleneck here is downstream of a measurement nobody has repeated, which is why section 7 treats one replication as worth more than further theory.

Frontier This subject depends on almost nothing exotic, which is unusual for an energy-systems entry and is the strongest practical argument for testing it. The apparatus is a machined ferrite shell, two electrodes, three multimeters and a thermally quiet room: no cryogenics, no vacuum, no fabrication capability a university machine shop lacks.

Established The first real dependency is materials metrology, because the prediction is only as good as mu-r and sigma. A replication that measures the permeability and conductivity of its own cylinder at the temperatures of the run is the difference between comparing two voltages and comparing two predictions. A traceable permeability measurement on a high-mu ferrite to better than five per cent is a national-metrology-institute job rather than a physics-department one, and Metrology Infrastructure covers why that distinction bites.

Established The second is a geomagnetic model good enough to give the axisymmetric field at each test site. The quantity in the theory is not the total local field an ordinary magnetometer reads but the sum of the m = 0 spherical-harmonic components, which comes from a global field model. Getting it wrong at a second site would produce a disagreement that looked like a failed replication and was not.

Frontier The third is low-noise DC measurement at the hundred-nanovolt level with orientation control — solved in principle, a craft problem in practice. What has to be engineered is a mount that rotates a 30 cm cylinder to four orientations without changing its thermal environment, and cabling whose junction thermoelectrics do not move when it does. Precision Quantum Sensing is the adjacent capability base.

Frontier The fourth is theoretical and is being worked: an agreed treatment of electromagnetic boundary conditions at a moving, permeable, conducting surface. That is the explicit content of the Brevik-Chaichian-Katsnelson Comment, a century-old corner of classical electrodynamics rather than new physics, and the kind of dispute settled by two groups agreeing on a worked example.

Speculative The fifth matters only if the first four resolve favourably: soft magnetic materials with a higher permeability-conductivity product that still sit below the Reynolds ceiling at useful sizes. That is a search over a well-characterised class, and Quantum Materials describes the shape of such searches and their usual disappointments.

Established The sixth is methodological rather than technical: preregistration, blinding and randomisation of orientation order. None was used in the original, none costs money, and all three are the standard response to a result whose plausible failure mode is an analysis choice. Forecasting Replication reports aggregated human forecasts pricing replication outcomes at about seventy per cent accuracy before anyone runs them. Frontier What this subject does not depend on is any result elsewhere in this corpus: nothing on the map blocks the decisive test, which sharpens the institutional question rather than the scientific one.

Frontier The decisive experiment is an independent replication at a different latitude, with co-located dummy cylinders recorded simultaneously and a length-scaling arm, and it is the test the proposers themselves name. Chyba’s formulation is that “some independent group needs to reproduce — or rebut — our results, with a system closely similar to our own”. This brief adds the second latitude, the simultaneous dummy and the scaling arm, because a same-latitude repeat that succeeds leaves the thermoelectric explanation exactly where it is, and one that fails is attributable to a hundred differences in workmanship.

Speculative The latitude arm is decisive because the published equations imply a steep latitude law that no artefact shares. The predicted emf goes as local surface speed times the square of the magnetic Reynolds number, and that Reynolds number is itself linear in surface speed, so the leading dependence is cubic in v and linear in the axisymmetric field. Surface speed falls as the cosine of latitude; the axisymmetric field strengthens poleward roughly as the square root of one plus three times the sine squared of latitude. Frontier Run against the Princeton values that gives about 24 µV at the equator, about 15 µV at Ottawa and about 1.3 µV at Resolute Bay — an eighteen-fold span, with the high-latitude point below the original measurement’s own uncertainty. A thermoelectric or instrumental artefact has no reason to vary by a factor of eighteen with latitude, and none to vary in the direction the theory picks.

Speculative Canada is the natural host for that arm, for reasons of geography rather than patriotism. One country spanning roughly 42 to 83 degrees north, with a long-running geomagnetic observatory network, can run one apparatus and one team at three latitudes the theory says should differ by more than an order of magnitude — cleaner than three groups in three countries, because it holds workmanship fixed.

Frontier The co-located dummy closes the thermal loophole, and it is the control the original lacks. The 2025 paper ran its solid-cylinder control as a separate campaign. Running a solid cylinder of the same material, outer dimensions and electrode arrangement on an identical channel, in the same mount, at the same time, makes the thermal environment common-mode. If both channels move together it is the room; if only the shell moves it is not.

Frontier The cheapest single upgrade is to stop measuring DC and modulate instead. Rotate the shell continuously and slowly about the local vertical — a revolution an hour is fast against any thermal time constant worth worrying about — and the claimed effect becomes a sinusoid at a known frequency with a known phase relative to magnetic north, while thermal drift stays at DC and in the 1/f tail. Speculative Lock-in detection at a mechanically enforced frequency gives a noise floor improving as the square root of integration time, and it requires a turntable.

Frontier The scaling arm is two cylinders and costs about $800. The predicted emf is linear in shell length, so a 60 cm cylinder should give twice the signal of a 30 cm one; a thermoelectric offset scales with the end-to-end temperature difference, not with length, and a longer cylinder in the same room generally has a smaller gradient per unit length. The two hypotheses predict opposite trends.

Speculative The hemisphere test is the most elegant and the most expensive. The same device in the same physical orientation should produce the opposite sign in Chile or New Zealand, because the axisymmetric field points out of the ground there. No thermal, capacitive or instrumental artefact named by the authors’ critics has a hemisphere.

Frontier The magnetic-shield null is the fastest sanity check. Enclose the assembly in a mu-metal can and the claimed effect must vanish while the Seebeck coefficient, the junction thermoelectrics and every instrumental offset stay where they were. The paper already has the converse result — a high-Reynolds-number MuMetal shell gave “emf = 0.019 ± 0.099 µV” — and the shielded-shell version is a day of work.

Frontier The material series converts a point into a curve. Nickel-zinc ferrites have conductivities orders of magnitude below manganese-zinc, so their predicted emf should collapse, while a higher permeability-conductivity product should raise it until the ceiling. Four grades on one rig tests the mechanism harder than any high-precision measurement of one.

Established The theory experiment is a calculation and is already half-run. Electromagnetic boundary conditions at a moving permeable surface is a well-posed problem with a determinate answer; Brevik, Chaichian and Katsnelson claim that imposing them correctly makes “the total power given to the shell... zero”. That is checkable by anyone with the patience and needs no hardware.

Handwave The experiment that would settle energy provenance does not exist and cannot be built at this scale. Measuring the retarding force directly means resolving three tenths of a femtonewton on a room-temperature ferrite cylinder. Nobody has proposed a way, and proposals that assume it away are where this subject stops being experimental.

Frontier Horizon and cost: every experiment above except the hemisphere arm fits inside a year and a budget under $50,000, and none is scheduled. The original apparatus cost about $25,000 to develop and the lead author puts a reproduction at under $10,000. Eighteen months after publication in a journal that featured the result, no group had announced an attempt.

Established There is no engineering problem in building the device; the engineering problem is measuring what comes out of it. The object is a passive lump of sintered ferrite with two electrodes. The instrumentation is the build: a mount that rotates without changing the thermal boundary conditions, low-thermal-emf cabling, temperature sensing at both ends to better than ten millikelvin, and a voltmeter whose own drift is below a hundred nanovolts over ten hours.

Frontier If the effect is real, the device is a very high impedance, very low voltage DC source, and that dictates the topology of anything built from it. At 681 ohms and 17.3 µV per unit, a one-volt supply is about 58,000 units in series carrying about 40 megohms of internal resistance — a source that can drive an electrometer and nothing else. Even at the theoretical Reynolds ceiling of roughly 220 µV per unit the string is still about 4,500 units. Frontier Contact offsets, not conversion efficiency, would be the limiting problem: the published apparatus carries a 1.20 ± 0.35 µV systematic on polarity reversal, which in a 58,000-unit string cancels to a few hundred microvolts if random and sums to tens of millivolts if systematic, and there is no published evidence about which.

Handwave Any engineering plan that assumes materials better than the ones that exist is not an engineering plan. The paper’s scaling sentence is conditional on materials “were materials with such characteristics available or created” — the correct way to say it, and an admission that the route runs through a discovery nobody has made.

Speculative The one engineering direction with real physics behind it moves the device off the rotating frame, and then it is a different subject. The paper names an orbiting satellite as a way to raise v; but a conductor in low Earth orbit is not corotating with the axisymmetric field, so ordinary motional induction applies without any loophole. That is the electrodynamic tether case, and Electrodynamic Propulsion Concepts covers the family of space systems that already exploit it. Invoking orbit concedes that the interesting claim is specifically about the ground.

Frontier The corpus holds two close analogues and they point the same way. Rotational Superradiance is the other rotation-extraction subject here with positive laboratory results — amplification measured in water, sound, light and an electromagnetic circuit — and it has the same unresolved question, whether the surplus energy came from the rotor or from the drive. Vacuum Energy Engineering is the corpus’s reference case for balance-sheet discipline. Established Both reach the same methodological conclusion from different physics: measuring an effect and demonstrating its source are separate experiments, and only the first has been done. Nothing here requires new infrastructure, which is why the subject belongs in the replication economy rather than the capital-programme one that Extreme-Environment Test Infrastructure describes.

Established The nearest adjacent technology is the one that makes the measurement hard: thermoelectric generation in ferrites. The same −417 µV K⁻¹ that contaminates this experiment is a respectable thermoelectric coefficient, and a MnZn ferrite cylinder across a real temperature difference is a functioning, if poor, thermoelectric generator. Any replication team becomes thermoelectric metrologists first and geomagnetic physicists second.

Established Motional induction in the geomagnetic field is routine engineering as soon as the conductor stops corotating. Electrodynamic tethers in low Earth orbit and the voltages induced in submarine cables by tidal currents are uncontroversial applications of v × B in Earth’s field. What makes this subject exotic is the corotation, not the induction. Established Unipolar induction is the classical ancestor of the dispute: whether the field of a rotating magnet rotates with it has generated a small permanent literature since the nineteenth century, and the frame-of-reference errors it produces are exactly the class Jeener alleges.

Frontier Geomagnetic observatory networks are the adjacent infrastructure and are already in place — continuous absolute magnetometry at fixed sites, coordinated internationally, which is what would supply the axisymmetric field values a multi-latitude replication needs, for reasons unrelated to this claim.

Frontier On the claims side the adjacent family is the carefully-bookkept energy propositions, not the free-energy ones. Quantum Energy Teleportation is the cleanest case: not an energy source, because a measurement at the sender injects the energy first. Quantum Batteries is a claim about rate, not capacity, with stored energy a six-thousandth of lithium. Established Both earn their place by stating what the accounting permits, which is the treatment this subject needs applied to a number that does not yet exist.

Frontier On the sociology side the adjacent cases tell you what happens next. Low Energy Nuclear Reactions describes an anomaly that became a funded, null-publishing programme, the best available outcome for a claim of this shape. Pathological Science is the natural history of the other outcome, and records that most such episodes die within weeks to a decade once someone blinds the observer. Speculative Neither has happened here, because the blinding experiment has not been run.

Frontier The energy-systems neighbours are useful mainly for scale. Wave and Tidal Energy already withdraws from the same planetary reservoir by a different mechanism, and has about half a gigawatt installed after fifty years and several billion dollars. Planetary Scale Energy Systems makes the point that transport, not capture, is usually the bottleneck — not the bottleneck here, because there is nothing yet to transport.

Established The institutional fact of this subject is that the decisive experiment costs less than a conference trip for four and nobody has run it. Under $10,000, months of work, no specialist facility, and a published, open-access, fully specified apparatus to copy. Eighteen months after publication the lead author was aware of no group attempting it.

Frontier The reason is structural rather than scientific: there is no funding line and no career return for replicating somebody else’s microvolt. A successful replication makes another group famous; a failed one makes enemies; neither wins a grant on its own. That is the standard replication-economics problem, biting hardest where it should bite least — on cheap, decisive, well-specified tests.

Frontier The institutions best placed to do it are national metrology institutes, not physics departments. The hard part is a stable microvolt DC measurement with orientation control and millikelvin thermometry, the core competence of the organisations that maintain voltage and temperature standards and something university groups must build from scratch. Metrology Infrastructure sets out why that capability is concentrated and rarely spent adjudicating other people’s claims.

Speculative Canada is unusually well placed to own the latitude arm, on the strength of a national asset rather than an ambition. A country spanning roughly 42 to 83 degrees north, with an established geomagnetic observatory network and a national metrology institute, can execute the one experiment that distinguishes a geomagnetic effect from a laboratory artefact using one apparatus and one team. No other jurisdiction gets that span inside one set of laboratory standards.

Established The journal machinery has, for once, worked as designed. The 2016 theory drew a published Comment and a published Reply in the same journal. The 2025 measurement spent twenty-one months in review, appeared open access, was flagged for the society’s magazine with external sceptics quoted in the same article, and drew a preprint Comment within six months. Frontier What the machinery does not do is commission the experiment, and no journal exists that could.

Speculative The institutional form this subject needs exists elsewhere in the corpus. Low Energy Nuclear Reactions documents a funded multi-laboratory programme whose explicit output includes nulls — a structure that pays for the answer rather than the hoped-for answer. A standing fund of a few hundred thousand dollars a year for replicating cheap high-leverage anomalies would clear this subject in a year and several like it. Frontier Cloud Laboratories is the other route, and records that the reproducibility argument for programmable laboratories is an argument from mechanism nobody has tested; this experiment is close to an ideal test case for it. Established And a public, dated, preregistered forecast on the replication costs nothing, as Forecasting Replication argues.

Established The authors behaved well and the coverage did not, and separating those is the first obligation here. The paper reports its own unexplained systematic, publishes the null controls that could have sunk it, and its lead author volunteered that the result “has a whiff of a perpetual motion machine” and that “it is very easy to make mistakes” at these voltages. That is the conduct a contested claim is supposed to attract.

Established The secondary coverage produced headlines the paper cannot support, including “Scientists Just Hacked the Planet... Could End the Fossil Fuel Era”. The same article concedes internally that “the power produced is minimal” and that results “need to be reproduced by other teams” — the standard structure of a headline that cannot be defended above a body that quietly withdraws it.

Frontier A microvolt result with a rotating-Earth story attached is unusually good raw material for investment fraud, and that is a foreseeable harm. The free-energy sector has a long record of attaching itself to real, cautiously-stated laboratory anomalies, and the features that make this claim scientifically interesting — a journal paper, a named university, an apparently impossible result — are exactly the features a prospectus needs.

Established The opposite error has a cost too. If the measurement is right it is a significant result in classical electrodynamics, and the cheapest way to find out has been available for eighteen months. Reflexive dismissal of a well-controlled measurement because its conclusion is unwelcome is not scepticism; it is the same epistemic move as credulity, run backwards. Frontier The honest line for public communication is that “an unreplicated microvolt” is the whole story and is interesting enough: every accurate account contains the words “not yet replicated” and every inaccurate one omits them, which is an unusually clean test for a reader.

Speculative If the effect were ever scaled, the resource question becomes a governance question with no existing owner. Nobody holds title to Earth’s rotation, and at any plausible scale the effect on the length of day is far below natural variability, so the problem is about claims and rents rather than physical harm — more tractable than the ones in Geoengineering Governance. Speculative The distributional shape would invert the fossil map, since the effect is predicted strongest near the equator and weakest near the poles. Handwave Drawing any policy conclusion from that today requires assuming the effect exists, scales, and is worth siting for — three assumptions this brief declines to make.

Speculative The civilizational case, stated conditionally because it has to be: a fuel-free, weather-independent, siting-flexible DC source with no thermal cycle and no moving parts would be a different kind of energy system from anything else in this corpus. No fuel chain to interdict, no intermittency to firm, no combustion products, no rotating machinery.

Established The resource is finite, shared and already being spent. Earth’s rotational kinetic energy is about 2.14 × 10²⁹ J; tidal friction dissipates roughly 3.7 TW of it, about ninety-five per cent in the oceans and five per cent in the solid Earth, lengthening the day by something like 2.5 ms per century while pushing the Moon out 3.8 cm a year.

Frontier A civilisation drawing 3.7 TW from rotation would roughly double the braking rate, taking the length of day from about 2.5 to about 5 ms per century. That is measurable — Earth-rotation services already resolve annual and semiannual length-of-day terms of 0.34 and 0.29 ms — and consequentially trivial. Any resource constraint on this technology would be economic and geometric, not planetary.

Handwave The geometric constraint is the one that bites, and at the demonstrated power density it is disqualifying. One gigawatt is a cube of ferrite about 950 km on a side; relaxing that requires a scaling law nobody has published, and the Reynolds ceiling caps the obvious route at about a factor of thirteen in voltage.

Speculative If the effect is real and does not scale, the consequence is scientific rather than energetic, and it is not small. A demonstrated exception to a two-century-old impossibility argument would send people back through the corner of classical electrodynamics where moving permeable media meet boundary conditions — the precise corner the published Comment says was mishandled — and that corner touches geodynamo modelling, magnetohydrodynamic metrology and the unipolar-induction puzzles.

Frontier The most likely outcome is nothing, and the corpus has the base rate. Anomaly Resolution in Physics counts zero confirmed new-physics outcomes across six high-profile discrepancies at three sigma or more since 2010. A single-group microvolt result with a live conventional explanation sits well inside that reference class. Established The institutional lesson survives either way: a published, fully specified, ten-thousand-dollar experiment that decides a contested question and goes unperformed for eighteen months says something about how research capacity is allocated that is true whether the effect is real or not.

These horizons track a replication programme and a theoretical dispute, not a technology roadmap, because there is no device here to develop — only a measurement to confirm or kill.

• 10 yr: Frontier Expect the boundary-condition dispute to be settled on paper well inside this horizon; it is a determinate problem in classical electrodynamics and three groups are already working it. Frontier Expect at least one independent replication attempt, most plausibly from a national metrology institute or a university group with existing nanovolt capability, and expect its result to be reported as the headline finding of the subject either way. Speculative A multi-latitude arm, which is the test that distinguishes a geomagnetic effect from an apparatus artefact, is plausible at this horizon and is not currently funded anywhere. Speculative If the replication fails, expect the subject to close the way Pathological Science says most such episodes close, within a few years of the first blinded test.
• 25 yr: Speculative If replication succeeds, the subject becomes a materials and scaling programme, and the Reynolds ceiling decides how long it lasts: a factor of about thirteen in voltage from the known route is not a power technology, so the programme either finds a new geometry or ends as a chapter in electrodynamics. Speculative Expect the result, if it survives, to be absorbed into precision magnetometry and geodynamo instrumentation long before it is absorbed into energy engineering. Handwave Anyone selling a rotation-powered generator at this horizon is selling the story, not the physics.
• 50 yr: Handwave A power-relevant version of this effect at this horizon requires a material with a permeability-conductivity product far beyond anything now known that still holds a magnetic Reynolds number below unity at usable dimensions, and no such search programme exists. Speculative The more likely fifty-year descendant is a class of geomagnetically-referenced DC sensors — a modest outcome traceable to the same equations.
• 100 / 250+ yr: Handwave Planetary-scale extraction from rotation at terawatt rates is a coherent thermodynamic idea with no known implementation path, and any timeline for it is a statement about the writer rather than about the technology. Handwave The nearest honest comparison in this corpus is Rotational Superradiance, where a measured amplification has still not been shown to draw on the rotor, and Vacuum Energy Engineering, where the accounting question has been open for seventy years.

• Depends on Nothing on this map blocks the decisive experiment, and that is the unusual fact about this subject: the apparatus is a machined ferrite shell and three voltmeters, and no capability another brief must first produce stands in the way. Where it leans on the map is for method and for scale-setting: Metrology Infrastructure supplies the traceable microvolt and millikelvin measurement the claim lives or dies by, Precision Quantum Sensing the low-noise measurement culture, and Anomaly Resolution in Physics the base rate against which a single unreplicated result should be read.
• Requires (not on this map) Everything here waits first on an independent replication of the ferrite-shell electromotive force at a second latitude, the one result that would move the assessment. The theory side waits on an agreed treatment of the electromagnetic boundary conditions at a moving permeable surface, the point on which the published Comment and the original derivation disagree. Comparing two measurements at all requires ferrite permeability and conductivity traceable to better than five per cent, because a thirty per cent permeability uncertainty enters the prediction squared. Any scaling programme would require soft magnetic materials with a higher permeability-conductivity product that still hold a magnetic Reynolds number below unity at usable dimensions, and none is known. The replication waits on a standing fund for replicating cheap decisive anomalies, since at under $10,000 the obstacle is not cost. The scaling and length tests need custom sintered ferrite shells in matched geometric series, a small-batch specialist order rather than a catalogue item.
• Enables If replication confirms the effect, this brief becomes the anchor for a class of geomagnetically-referenced DC sensors and for a reopening of the moving-media corner of classical electrodynamics that Quantum Materials and geodynamo modelling both touch. If replication fails, it enables something almost as useful: a worked public example of a cheap decisive test closing a contested claim, which is the outcome Low Energy Nuclear Reactions shows is achievable when somebody pays for it.
• Adjacent Rotational Superradiance is the closest sibling — the corpus’s other rotation-extraction subject with positive laboratory results and an unresolved energy provenance; Vacuum Energy Engineering supplies the balance-sheet discipline; Quantum Energy Teleportation and Quantum Batteries are the corpus’s other carefully-bookkept energy claims; Electrodynamic Propulsion Concepts is where uncontroversial motional induction in Earth’s field already does work; and Wave and Tidal Energy is the only technology that already withdraws from the same reservoir at scale.

Handwave “This could end the fossil fuel era.” A real headline about this paper. The measured device delivers about 1.1 × 10⁻¹³ W from 9.4 × 10⁻⁵ m³; a gigawatt at that density is a cube of ferrite roughly 950 km on a side; and the paper declines to estimate scaling, saying it “will be the subject of subsequent investigations”. Established No published document, the paper included, claims a path to useful power.

Established “It is a perpetual motion machine.” No — the proposed mechanism is a brake. Earth’s rotational energy is a finite store of about 2.14 × 10²⁹ J that tidal friction already drains at about 3.7 TW; a device drawing on it would slow the planet, which is thermodynamically ordinary. Frontier The smell comes from elsewhere — the impossibility argument says the emf cancels, not that energy would come from nowhere.

Frontier “Energy conservation rules it out.” Not at this scale, and the reason is arithmetic. The retarding force needed to supply 1.1 × 10⁻¹³ W at 354 m s⁻¹ is 3 × 10⁻¹⁶ N, undetectable on this apparatus by any proposed means. Established Energy conservation is not in dispute between the parties; a boundary condition is.

Established “Researchers disprove their own work.” A widely-syndicated headline that inverts the sequence. The impossibility proof is the standard textbook one, not the authors’. In 2016 Chyba and Hand identified an assumption inside it and argued it can be violated; the 2025 paper tests that argument. They proposed a loophole in somebody else’s theorem and then measured.

Frontier “Faraday tried this in 1832 and failed, so it is settled.” Faraday’s attempt to detect a motional electromotive force in Earth’s field is the canonical origin story and is invoked in most coverage of this result. It is also a different configuration, and a nineteenth-century null on different apparatus has no bearing on a claim about magnetically permeable shells.

Frontier “The 2018 null already settled it.” It settled a different experiment. Veltkamp and Wijngaarden measured “more than 3 orders of magnitude smaller than predicted” on an antenna that, on the 2025 paper’s account, did not satisfy the topological and low-Reynolds-number conditions. Speculative Whether that is a fair distinction or a retreat is what a replication of the ferrite geometry would decide, and it is why the geometry must be copied rather than reinvented.

Established “It has already been refuted.” The published refutations target the theory, not the measurement. Jeener’s Comment refutes the 2016 derivation; the Brevik-Chaichian-Katsnelson Comment re-derives the same problem with different boundary conditions. Frontier Neither accounts for what produced 17.3 ± 1.5 µV with the reported orientation dependence, and a theory objection that leaves the data unexplained is half an answer.

Frontier “It is obviously the Seebeck effect.” It might be, and the paper is not naive about it. The authors measured the coefficient themselves at −417 µV K⁻¹, instrumented both ends to 0.01 °C, and separated the contributions by signature rather than subtraction. Frontier What keeps the explanation live is that 41 mK of steady gradient reproduces the whole signal, orientations were set by physically rotating the apparatus and then waiting overnight, the order was not randomised, and a 1.20 ± 0.35 µV polarity-reversal systematic remains unexplained. That is a live alternative, not a refutation.

Established “They replicated it at a second site.” They controlled for site noise at a second site. The second location was “a residential building 5.5 km east” — same latitude, same field, same cylinder, same team. A good control, not an independent replication, and the paper does not claim it is one.

Handwave “Scaling is just engineering.” No, because the mechanism carries its own ceiling. The effect is claimed only below a magnetic Reynolds number of one and the predicted emf rises as its square, so the entire headroom of the known route is the gap between 0.088 and 1 — about a factor of thirteen in voltage once the logarithmic term is included. Speculative Beyond that requires a material nobody has made.

Frontier “Earth’s rotational energy explains tectonics and earthquakes.” A separate family of claims travels alongside this one. A 2025 preprint argues that the kinetic energy change from periodic variations in Earth’s rotation rate — about 6.8 × 10²⁰ J per day, or 2,794.5 Tsar Bombas — “is sufficient to fully sustain all tectonic activity”, and grounds an earthquake-prediction method on it. Handwave It is not peer reviewed, and its author offers a Eurasian patent grant as evidence of examination, which is a finding about novelty rather than correctness. Nothing in the Chyba-Hand line supports or requires it.

Established “Nobody serious takes this seriously.” The record says otherwise, and so does the disagreement. The measurement was published in a Physical Review journal after twenty-one months of review, featured by the society’s magazine, and covered by Nature under the headline “Physicists divided over controversial claim”. Rovelli’s response was “maybe there is a subtler version of the argument that rules out this possibility; I do not know”. Frontier Division is the accurate description; dismissal is not, and neither is endorsement.