A Difference of Large Numbers

What Problem Are We Actually Talking About?

Since 1989, a contested area of research has asked whether small metal electrodes — typically palladium — can produce more heat than the electricity flowing through them. The field has gone by several names: cold fusion, low-energy nuclear reactions (LENR), condensed matter nuclear science. The name that has stuck in technical circles is solid-state fusion, or SSF.

The central claim sounds simple: put electrical power in, measure thermal power out, and if the output exceeds the input by more than any chemistry can account for, something interesting is happening. But excess heat is never measured directly, the way you’d read a voltage from a meter. It is a residual — the number left after subtracting everything else from an energy balance. And it sits between quantities that are each individually large.

That structure makes it a measurement problem first, and a physics problem second. This article is about the measurement problem.

The Energy Balance: What Goes In and Out of an Electrochemical Cell

Think through what is physically happening in a typical SSF experiment. A palladium electrode sits in heavy water (water made with deuterium, D₂O, in place of ordinary H₂O) and a current is run through it. The goal is to load enough deuterium into the palladium crystal lattice that something anomalous might occur. Measuring whether it does requires tracking every joule entering and leaving the cell.

Electrical power in — V × I — is easy to measure well. But it doesn’t all become heat. A significant share drives the electrolysis reaction: splitting D₂O into deuterium gas and oxygen. Splitting water is endothermic — it consumes energy. The energy per unit charge required to drive this reaction is fixed by a quantity called the thermoneutral potential, roughly 1.5 V for this system.¹ Any current multiplied by that voltage goes into chemistry, not heat, and has to be subtracted before any thermal surplus can be claimed.

Then come the other correction terms. Some of the deuterium and oxygen that evolve recombine back into water inside the cell and release their chemical energy as heat — this must be added back in. Water vapor escapes and carries enthalpy away with it. Heat escapes through the cell walls, the wires, the electrodes — by conduction, convection, and radiation — each with its own temperature dependence. As the water level drops over the course of a run, the thermal geometry of the cell changes, which changes how it loses heat.

Excess heat, if it exists at all, is what remains after accounting for all of that: measured output minus input minus chemistry minus losses. When the claimed surplus is one to two percent of the total input — say, 100 to 300 milliwatts on a 10-watt cell — every single term has to be known to better than one percent, and that accuracy has to be maintained over days or weeks, the time it takes the palladium to absorb enough deuterium to be relevant.

This is, by any measure, a very hard energy balance to close. The difficulty is structural, not incidental: you are taking a small difference between large, individually uncertain quantities, and holding that difference steady while the apparatus slowly changes underneath you.

What a Careful Null Result Teaches

The most instructive single result in the recent literature isn’t a positive finding. In 2016, a team at ReResearch LLC tried to replicate a published excess-heat experiment over 231 trials. They measured an average excess power of 6.1 ± 21.6 milliwatts on a cell drawing about 10 watts of input.²

Read those numbers carefully. The claimed effect in the original experiment ran to roughly 100 to 300 milliwatts. The replication team couldn’t see it. But the reason was not primarily statistical noise. When they characterized their hardware, they found:

• Calibration drift over the run: 130 to 460 milliwatts
• Thermistor shifts from mechanical instability of the cell cap: 135 to 324 milliwatts
• A single sensor excursion: ~400 milliwatts

Each artifact was as large as, or larger than, the effect being sought. The signal wasn’t lost in the arithmetic. It was lost in the physical noise floor of the apparatus.

This is an important distinction for anyone trained in measurement. You cannot fix a hardware noise floor by doing better statistics. If your calibration drifts by hundreds of milliwatts on a 10-watt system, no analysis rescues the residual. The floor is set by the front end — the sensors, the thermal design, the reference stability — and lowering it is a hardware and instrumentation problem, not a computation problem.

The Calibration Argument — and Why It Can’t Be Settled by Argument

The subtlest challenge to the SSF excess-heat data concerns calorimeter stability. Physicist Kirk Shanahan proposed that a shift in a cell’s effective calibration constant — caused by heat generation moving around inside the cell as the experiment runs — could produce an apparent surplus of roughly the right size, with no real anomaly at all.³ A ten-author team replied that the calibration data don’t support a shift that large, and that the proposed mechanism doesn’t survive in designs where the heat is captured outside the cell.⁴

One doesn’t need to pick a side to see the shape of the exchange. Both sides are arguing in calorimetry’s own terms. The dispute will be closed by better calorimeters, not by better arguments — which is the correct outcome for a measurement question.

The Calorimetry Toolkit: Why Using More Than One Method Matters

Precision calorimetry isn’t one technique — it’s a family of them, and applying more than one to the same experiment is the field’s strongest available tool, because different methods carry different systematic errors.

Isoperibolic calorimetry (used in the original Fleischmann-Pons experiments) holds the surroundings at a constant temperature and infers heat output from calibration curves. It is sensitive but depends entirely on the calibration being correct and stable throughout a multi-day run.⁵

Flow calorimetry circulates a coolant past the cell and measures its temperature rise. Total heat output is mass flow rate × specific heat × ΔT — a standard approach in chemical engineering for measuring heat from reactions. It gives an independent number with a different set of potential systematic errors.⁶

Seebeck calorimetry surrounds the cell with thermoelectric elements — the same principle as a thermocouple — that convert heat flux directly into a voltage. It integrates all the heat leaving the cell regardless of where inside the cell it originates, which matters if the heat source is spatially localized.⁷

If two or three of these methods agree on the same cell, the result is far more credible than any single method can provide, because their systematic errors don’t overlap. This kind of cross-validation has rarely been applied systematically in SSF work, and its absence is one of the field’s most significant methodological gaps.

The Nuclear Physics Constraint: What Calorimetry Alone Cannot Answer

Even if the energy balance were closed perfectly and a genuine surplus survived every correction, there is a separate problem that calorimetry cannot solve on its own.

If the excess heat is nuclear in origin, nuclear physics makes very specific predictions about what else must be produced. One watt sustained by deuterium fusion — deuterium being the hydrogen isotope with one neutron — would require roughly 260 billion reactions per second, each releasing about 23.8 million electron-volts of energy.⁸ At that rate, the two ordinary deuterium fusion pathways would produce large fluxes of neutrons and tritium (a radioactive form of hydrogen). An alternative pathway that produces helium-4 directly would instead emit 23.8 MeV gamma rays — extraordinarily energetic radiation that detectors would catch immediately. Neither signature has been reliably observed at the levels the claimed heat would require.⁹

That absence is the central unanswered scientific objection to interpreting the heat as nuclear in origin. It doesn’t prove that nothing unusual is happening. But it means that a heat measurement, however well-done, is not by itself evidence of nuclear reactions. Connecting the two requires measuring both heat and any reaction byproducts simultaneously, on the same clock.

Why This Is Relevant Beyond Cold Fusion

You might ask why anyone without a stake in the cold fusion controversy should care about this. The answer is that SSF hands precision calorimetry an unusually demanding test case — one that can sharpen the discipline.

The requirements — micro- to milliwatt residuals on multi-watt inputs, baselines that must stay stable for weeks, a sample that changes its own properties while the experiment runs — are exactly the conditions that separate good instruments from great ones. The techniques for handling them (synchronous and ratiometric measurement, careful treatment of thermocouple cold junctions, calibration checked under actual working conditions rather than assumed to persist) transfer directly to other measurement problems: tracking slow heat release in aging batteries, detecting faint signals from catalytic reactions, or mapping hydrogen behavior in metal storage systems.

In 2019, a team convened by Google published in Nature a carefully done negative result across 420 samples — no excess heat — and then did something useful: they named exactly where the older work had been too crude to say anything definitive. The materials science of extreme hydrogen loading, and the thermal measurements built around it, remained underexplored with modern tools.¹⁰ In 2023, ARPA-E — the U.S. government’s advanced energy research agency — committed roughly $10 million across eight teams, including groups at MIT, Stanford, and Lawrence Berkeley National Laboratory, to test LENR claims with modern instrumentation.¹¹

That is an invitation to do better measurement, regardless of what the answer turns out to be.

What Would a Decisive Measurement Actually Require?

The honest summary: the question of whether anomalous excess heat exists in SSF experiments is unanswered, not because the answer is known and concealed, but because the measurements so far haven’t been good enough to answer it cleanly.

A null result — no excess heat found — only means something if the instrument was sensitive enough to have detected the claimed effect. When the measurement uncertainty is larger than the claimed signal, a null result tells you nothing. It is silence, not absence.

Lower the hardware noise floor by an order of magnitude and the situation changes. A surplus that survives careful, cross-validated scrutiny at that sensitivity is a real anomaly — something the current picture can’t explain, and worth understanding. A clean null at sufficient sensitivity finally closes a chapter that has been open since 1989. Both outcomes are useful.

The only worthless outcome is the one the field has produced too often: a number no one outside the room can check.

Is there anomalous excess heat in SSF experiments? It is a calorimetric question, and it has a calorimetric answer. The discipline has the instruments. What it has lacked is their patient, complete, cross-validated application to the one measurement that has waited more than thirty years for it.

Editorial note: This primer presents a scholarly synthesis of solid-state fusion's relationship to precision calorimetry, written for a college-level audience. The underlying nuclear claims of SSF/LENR remain scientifically contested. Readers are directed to primary experimental literature for empirical evaluation. Rev. 2026-08-29.

References & Footnotes

1. The fraction of electrical input consumed by the water-splitting reaction (rather than heating the cell) is fixed by the thermoneutral potential, approximately 1.48 V for light water and 1.53 V for heavy water at room temperature. This value comes from the enthalpy of water formation divided by the charge transferred per mole; it is standard electrochemical thermodynamics. ↩
2. M. J. Guffey, Y. Tang, and P. J. King (ReResearch LLC), “Attempted Replication of Excess Heat in the Letts Dual-Laser Experiment,” Journal of Condensed Matter Nuclear Science 20 (2016): 1–28. The team measured 6.1 ± 21.6 mW over 231 trials on a ~10 W cell, against an original claim of 100–300 mW excess, and itemized hardware artifacts each comparable to the effect being sought. ↩
3. K. L. Shanahan, “Comments on ‘A New Look at Low-Energy Nuclear Reaction Research,’” Journal of Environmental Monitoring 12 (2010): 1756–1764. Shanahan argues that a shift in a cell’s effective calibration constant — caused by the location of heat generation migrating inside the cell — can produce an apparent excess of the reported magnitude. ↩
4. J. Marwan et al. (ten authors), “A New Look at Low-Energy Nuclear Reaction (LENR) Research: A Response to Shanahan,” Journal of Environmental Monitoring 12 (2010): 1765–1770. ↩
5. M. Fleischmann, S. Pons, et al., “Calorimetry of the Palladium–Deuterium–Heavy Water System,” Journal of Electroanalytical Chemistry 287 (1990): 293–348. Isoperibolic calorimetry holds the surroundings at constant temperature and infers heat loss from calibration; see W. Hemminger and G. Höhne, Calorimetry: Fundamentals and Practice (Weinheim: Verlag Chemie, 1984) for method definitions. ↩
6. Flow (mass-flow) calorimetry is standard in chemical reaction engineering. M. C. H. McKubre et al. at SRI International built much of their excess-heat work on mass-flow methods; see McKubre, “Cold Fusion (LENR): One Perspective on the State of the Science,” Journal of Condensed Matter Nuclear Science 4 (2011): 32–44. ↩
7. Seebeck (heat-conduction) calorimetry uses thermoelectric elements surrounding the cell to integrate total heat flux. Hemminger and Höhne, Calorimetry: Fundamentals and Practice. ↩
8. One watt from D + D → ⁴He (Q = 23.8 MeV = 3.8 × 10⁻¹² J) requires about 2.6 × 10¹¹ reactions per second. Q-values and branching ratios from D. A. Brown et al., “ENDF/B-VIII.0,” Nuclear Data Sheets 148 (2018): 1–142. ↩
9. The expected reaction byproducts — neutrons, tritium, and/or high-energy gamma rays — have not reliably appeared at the levels the heat would demand. U.S. Department of Energy, Report of the Review of Low Energy Nuclear Reactions (December 2004); Edmund Storms, The Science of Low Energy Nuclear Reaction (Singapore: World Scientific, 2007). ↩
10. C. P. Berlinguette et al., “Revisiting the Cold Case of Cold Fusion,” Nature 570 (2019): 45–51. The Google-convened team found no excess heat across 420 samples but identified extreme hydrogen loading and the thermal measurements around it as underexplored. ↩
11. Advanced Research Projects Agency–Energy (ARPA-E), “U.S. Department of Energy Announces $10 Million in Funding to Projects Studying Low-Energy Nuclear Reactions,” February 2023. ↩

Introduction: A Small Number Living Inside a Large One

Ask a precision calorimetrist what they fear most and the answer is rarely a mysterious effect. It is a small number that lives inside a large one. Solid-state fusion (SSF), the field once called cold fusion and now more carefully named condensed matter nuclear science, or low-energy nuclear reactions, has spent more than thirty years disputing a quantity of exactly that kind. The excess heat everyone argues about is not a measurement in the way a voltage is a measurement. It is a residual in an energy balance: the thermal power leaving a cell, minus the electrical power going in, minus the chemistry, minus the losses, with the claim being that what is left over is larger than any known chemical source can supply.¹

That framing should interest a thermal-measurement specialist, because it locates the whole controversy on home ground. The disputed object is not, in the first instance, nuclear physics. It is an energy balance that has to close, and closure is the calorimetrist’s craft. Most of the famous fights over SSF have been closure fights in disguise. Were all the inputs counted? Was the calibration still good after three weeks of drift? Were the heat losses characterized, or assumed? These are not questions about the atom. They are questions about the instrument and the bookkeeping around it, and the discipline that answers them has advanced a long way since 1989.

The Residual Is a Difference of Large Numbers

Consider what a complete balance on an electrolytic SSF cell actually contains. Electrical power flows in, easily measured. Most of it does ordinary work: driving the endothermic electrolysis reaction and warming the cell. The share that splits water rather than heats it is fixed by the thermoneutral potential, roughly 1.5 volts for these systems, and it has to be subtracted before any surplus can be claimed.² Then come the corrections that sank more than one early result. Some of the evolved deuterium and oxygen recombine inside the cell and release their chemical energy back as heat. Water leaves as vapor, carrying enthalpy with it, and the level drops, changing the cell’s thermal behavior as it goes. Heat escapes by conduction, convection, and radiation, each with its own temperature dependence. The excess, if it exists, is the difference between the measured output and the sum of all of these.

When the surplus being sought is a percent or two of the input, every one of those terms has to be known to a fraction of a percent, sustained over the days or weeks a loaded cathode takes to do anything. This is the structural reason the field has been so hard to settle. It is not that excess heat is too subtle to detect. It is that detecting it means winning a subtraction between quantities that are individually large and individually uncertain, and holding that win steady while the apparatus slowly changes underneath you. Fleischmann and Pons understood this, which is why their central papers were calorimetry papers, built on isoperibolic measurement in which the cell sits in constant-temperature surroundings and heat loss is inferred from calibration.³ The method was sound. The question that has never fully closed is whether their particular implementation of it accounted for everything it needed to.

What a Good Null Actually Teaches

The most instructive result in the modern literature is a negative one. In 2016 a group at ReResearch LLC set out to replicate a well-known dual-laser excess-heat claim and found nothing: 6.1 ± 21.6 milliwatts of excess power averaged over 231 trials, on a cell drawing close to 10 watts.⁴ Read that as a calorimetrist and the lesson jumps out. The claimed effect ran to a percent or two of the input, around 100 to 300 milliwatts; what the replication could actually resolve was a residual near a thousandth of the input, with an uncertainty almost four times any central value. More telling still, the confounds the authors itemized were not analysis errors. They were thermal hardware: calibration drift worth 130 to 460 milliwatts, thermistor shifts from mechanical instability of the cell cap worth 135 to 324 milliwatts, a single-sensor excursion near 400 milliwatts. Each of these was as large as, or larger than, the effect being sought. The floor that swallowed any possible signal was built of physical apparatus, not arithmetic.

That the floor is physical rather than arithmetic is not a worry invented for this essay. The most substantive published challenge to the excess-heat corpus is itself a calibration argument. Kirk Shanahan has contended that an unrecognized shift in a cell’s calibration constant, driven by the location of heat generation migrating inside an open cell as recombination moves to the electrode, can manufacture an apparent excess of about the right size.⁵ A ten-author group replied that the calibration data do not support a shift that large, and that in any case the mechanism cannot survive a move to closed-cell, mass-flow, or Seebeck designs that capture the heat outside the cell.⁶ One need not adjudicate that exchange to see its shape. It is a dispute conducted in calorimetry’s own terms, and it will be closed by better calorimeters rather than by argument.

That is the crux, and it cuts in a direction a careful practitioner should welcome. A better analysis cannot rescue a measurement whose baseline wanders by hundreds of milliwatts; that floor is set by the front end, the reference, and the thermal design of the calorimeter itself. Which is precisely why lowering it is a calorimetry problem rather than a debating problem. A small residual riding on a restless baseline is a familiar adversary in precision thermal work, answered with the ordinary tools of the trade: synchronous and ratiometric measurement, low-thermal-EMF connections, cold-junction compensation done properly, references traceable to a standard, and calibration checked against the working conditions rather than assumed to persist. A thermocouple, after all, is a microvolt sensor whose cold junction is itself a temperature measurement. Treat that reference casually and the drift you record is your own instrument breathing.

The payoff is specific. Lower the hardware floor by an order of magnitude and the 2016 null stops being ambiguous. A signal that was buried becomes one you can confirm or exclude. It is worth being exact about what that buys, though. A null retires a hypothesis only when the instrument’s sensitivity clears the effect size being claimed. At or below the error floor, a null is silence, not absence.⁷ The entire value of a better calorimeter is that it pushes the sensitivity to where the answer, either answer, finally carries information.

The Closure Toolkit

Modern calorimetry offers more than one road to the same number, and their independence is the point. Flow calorimetry, which reads the temperature rise of a coolant carrying heat away from the cell, gives a more complete capture of total output than an isoperibolic setup and suits higher-power experiments; it is standard in reaction engineering, and Michael McKubre’s group at SRI International built much of its excess-heat work on mass-flow methods of exactly this kind.⁸ Seebeck calorimetry surrounds the cell with thermoelectric elements that integrate all the heat leaving it, whatever its internal distribution, which matters when the source may be localized rather than uniform.⁹ Differential scanning calorimetry and the broader thermodynamics of metal hydrides supply the chemical half of the ledger: the enthalpies of hydride formation and decomposition, and the loading behavior that the resistance ratio of the cathode tracks up toward the D/Pd ≈ 0.85 to 0.90 range that excess-heat reports tend to cluster near.¹⁰ Run two or three of these methods on the same cell and agreement between them is worth far more than any single measurement, because their systematic errors do not overlap. Cross-validation of this kind is the strongest instrument the field has, and it has rarely been applied systematically across the corpus.

There is a hard limit that calorimetry alone cannot pass, and honesty about it strengthens rather than weakens the case for better measurement. Suppose the closure were perfect and a genuine surplus survived every correction. One watt sustained by deuterium fusing to helium-4 would require about 2.6 × 10¹¹ reactions per second, each releasing 23.8 MeV, and a rate that high cannot stay hidden.¹¹ The two ordinary deuteron-deuteron channels split almost evenly, and at that rate they would flood the cell with neutrons and tritium; the direct fusion-to-helium-4 route that the heat-helium reports implicitly invoke would instead emit a matching flux of 23.8 MeV gamma rays. Neither signature has reliably appeared alongside the reported heat, not the neutrons and tritium of the ordinary branches, not the hard gamma of the helium route, at anything near the level the heat would demand. That absence is the central unanswered objection to a nuclear interpretation, and it is an argument about counts and cross-sections, not metaphysics.¹²

It is worth naming the prior plainly. Thirty years of predominantly negative and non-replicated results, set against a claimed reaction pathway that sits athwart well-tested nuclear physics, put the probability of a genuine nuclear effect low.¹³ That is not a reason to stop measuring. It is precisely why the evidentiary bar is high and why a clean measurement, either way, would be worth so much.

The lesson for the discipline is therefore not discouraging. It is that calorimetry’s job is to establish the anomaly, cleanly and with stated uncertainty, and never to pronounce on its cause. A well-closed excess-heat measurement is a thermodynamic fact in need of an explanation. It is not, by itself, evidence that the explanation is nuclear. The measurement that would settle origin couples the calorimeter to a product detector on a shared clock, so that heat and ash, if any, rise together. Keeping those two questions apart is what makes the whole enterprise credible.

What the Home Field Gets Back

None of this is a favor to a fringe. SSF hands precision calorimetry an adversarial test case at the outer edge of what the discipline can do, and the return traffic is real.

The measurement is a worst case worth owning. Micro- to milliwatt residuals on multi-watt inputs; baselines that must hold steady for weeks rather than seconds; a source that may be spatially localized and time-varying; and a sample, the loaded cathode, that changes state unpredictably as it runs. Solve calorimetry-grade closure and long-baseline stability against that, and the methods travel directly to any field chasing a faint thermal signal against a shifting background, from heterogeneous catalysis to the slow self-heating of aging batteries. The thermodynamic side profits too. Pushing palladium and nickel hydrides to extreme loading forces measurements of hydrogen chemical potential, phase boundaries, and formation enthalpies in regimes the hydrogen-storage literature has not mapped well, and those data are useful wherever hydrogen-in-metal behavior matters. And the field’s need for shared calibration protocols, reference cells, and reporting standards is an invitation to write calorimetric best practice that would serve any precision thermal measurement, not only this one.

The institutional pull is already there. In 2023 the U.S. Advanced Research Projects Agency–Energy committed roughly $10 million across eight teams, including groups at MIT, Stanford, and Lawrence Berkeley, to test low-energy nuclear reaction claims with modern instrumentation.¹⁴ A few years earlier a Google-convened effort published a careful negative result in Nature, no excess heat across 420 samples, and, more usefully, named where the older work had been too crude to conclude anything: the materials science of extreme loading and the measurements built around it.¹⁵ That is a standing invitation to bring better thermal metrology to an under-measured problem, issued by people with no stake in the answer being positive.

The Balance Can Be Made to Close

The single most useful thing calorimetry can contribute to solid-state fusion is not a verdict. It is an energy balance closed well enough that its residual means something whichever way it falls. Full accounting of every input and loss term, several independent calorimetric methods run on the same cell, thermodynamic data good enough to fix the chemical baseline, and every result reported with the sensitivity that makes a null informative. Two of the three possible outcomes are valuable. A surplus that survives that scrutiny is a real anomaly and the start of a genuine problem for the standard picture. A clean null at a sensitivity that clears the claimed effect finally retires a stretch of the map that has stayed grey for a generation. The only worthless result is the one the field has produced too often: a number no one outside the room can check.

Is there anomalous excess heat in SSF experiments? It is a calorimetric question, and it has a calorimetric answer. The discipline has the instruments. What it has lacked is their patient, complete, cross-validated application to the one measurement that has waited thirty years for it.

Editorial note: This primer presents a scholarly synthesis of solid-state fusion's relationship to precision calorimetry, written for an expert audience. The underlying nuclear claims of SSF/LENR remain scientifically contested. Evidence tiers (A = replicated/consensus, B = single peer-reviewed source, C = preliminary/contested) are noted inline. Readers are directed to primary experimental literature for empirical evaluation.

References & Footnotes

1. Excess heat in SSF calorimetry is not read directly but computed as a residual in an energy balance, derived from raw temperature, current, and voltage records through a calorimetric model (baseline subtraction, calibration, loss model, and integration over time). Edmund Storms, The Science of Low Energy Nuclear Reaction (Singapore: World Scientific, 2007); standard calorimetric practice. The framework is established (tier A); the SSF calorimetric record specifically remains contested (tier C). ↩
2. The portion of electrical input that drives the endothermic water-splitting reaction, rather than heating the cell, is fixed by the thermoneutral (enthalpy-based) cell potential: about 1.48 V for light water and roughly 1.53 V for heavy water at room temperature, from ΔH/nF with ΔH the enthalpy of formation of the water. It must be subtracted before any thermal surplus is claimed. Standard electrochemical thermodynamics; see also the treatment of the thermoneutral potential for the Pd–D₂O system in the condensed-matter-nuclear-science calorimetry literature. Tier A. ↩
3. M. Fleischmann, S. Pons, M. W. Anderson, L. J. Li, and M. Hawkins, “Calorimetry of the Palladium–Deuterium–Heavy Water System,” Journal of Electroanalytical Chemistry 287 (1990): 293–348. Isoperibolic calorimetry holds the surroundings at constant temperature and infers heat loss from calibration; see W. Hemminger and G. Höhne, Calorimetry: Fundamentals and Practice (Weinheim: Verlag Chemie, 1984) for method definitions. Tier A for the method; the adequacy of any specific implementation’s energy accounting is the contested question. ↩
4. M. J. Guffey, Y. Tang, and P. J. King (ReResearch LLC, Hawthorne, CA), “Attempted Replication of Excess Heat in the Letts Dual-Laser Experiment,” Journal of Condensed Matter Nuclear Science 20 (2016): 1–28. Reported excess power 6.1 ± 21.6 mW over 231 laser-triggered trials at ~10 W input, against an original claim of 100–300 mW; itemized artifacts of hardware origin: calibration 130–460 mW, thermistor shifts from mechanical instability of the calorimeter cap 135–324 mW, plus a ~400 mW single-sensor excursion. Published negative replication; tier B. ↩
5. K. L. Shanahan, “Comments on ‘A New Look at Low-Energy Nuclear Reaction Research’,” Journal of Environmental Monitoring 12 (2010): 1756–1764; and “A Systematic Error in Mass Flow Calorimetry Demonstrated,” Thermochimica Acta 387 (2002): 95–100. Shanahan’s calibration-constant-shift (CCS) hypothesis holds that a change in a cell’s effective calibration constant, driven by the location of heat generation (including recombination at the electrode) migrating within an open cell, can produce an apparent excess of roughly the reported magnitude. Peer-reviewed critique; tier B, disputed. ↩
6. J. Marwan, M. C. H. McKubre, F. L. Tanzella, P. L. Hagelstein, M. H. Miles, M. R. Swartz, E. Storms, Y. Iwamura, P. A. Mosier-Boss, and L. P. G. Forsley, “A New Look at Low-Energy Nuclear Reaction (LENR) Research: A Response to Shanahan,” Journal of Environmental Monitoring 12 (2010): 1765–1770. The ten-author reply argues that the calibration data do not support a shift of the required size, and that the CCS mechanism cannot account for excess heat measured in closed-cell, mass-flow, or Seebeck configurations, where recombination and heat capture are handled differently. Peer-reviewed; tier B. ↩
7. A null result retires a hypothesis only if the measurement’s sensitivity clears the claimed effect size; at or below the error floor, a null is absence of evidence at insufficient power, not evidence of absence. Standard statistical-power and detection-limit reasoning. ↩
8. Flow (mass-flow) calorimetry measures the temperature rise of a flowing coolant to capture total heat output and provides an independent cross-check on isoperibolic results; it is standard in chemical reaction engineering. M. C. H. McKubre et al., “Mass Flow Calorimetry,” in the condensed-matter-nuclear-science literature (SRI International program); see also McKubre, “Cold Fusion (LENR): One Perspective on the State of the Science,” Journal of Condensed Matter Nuclear Science 4 (2011): 32–44. Tier A for the technique. ↩
9. Seebeck (heat-conduction) calorimetry uses thermoelectric elements surrounding the cell to integrate the total heat flux leaving it, largely independent of where inside the cell the heat is produced. Hemminger and Höhne, Calorimetry: Fundamentals and Practice. Tier A for the technique. ↩
10. Differential scanning calorimetry and metal-hydride thermodynamics supply hydride formation and decomposition enthalpies and heat capacities needed to fix the chemical baseline; the resistance ratio of the cathode is the standard non-destructive loading diagnostic, and excess-heat reports cluster at high loading, D/Pd ≈ 0.85–0.90, with high loading necessary but not sufficient. M. C. H. McKubre et al., “Using Resistivity to Measure H/Pd and D/Pd Loading: Method and Significance,” in Condensed Matter Nuclear Science (ICCF-12 Proceedings) (Singapore: World Scientific, 2006); Hemminger and Höhne for DSC. Tier B; threshold values approximate. ↩
11. One watt from D + D → ⁴He (Q = 23.8 MeV = 3.8 × 10⁻¹² J) requires about 2.6 × 10¹¹ reactions per second (arithmetic). The two principal deuteron–deuteron branches, t + p and ³He + n, occur with roughly equal probability and yield energetic tritium, protons, and neutrons; the direct ⁴He + γ branch is strongly suppressed (branching ~10⁻⁶–10⁻⁷) and would emit a 23.8 MeV gamma. Q-values and branching from D. A. Brown et al., “ENDF/B-VIII.0: The 8th Major Release of the Nuclear Reaction Data Library,” Nuclear Data Sheets 148 (2018): 1–142. Tier A. ↩
12. A commensurate flux of the expected products has not reliably appeared alongside the reported heat, which is the central unresolved objection to a nuclear interpretation. U.S. Department of Energy, Report of the Review of Low Energy Nuclear Reactions (Washington, DC, December 2004); Storms, The Science of Low Energy Nuclear Reaction. General absence tier A; specific positive product reports remain single-community and contested (tier C). The reported correlation of excess power with ⁴He (M. H. Miles, B. F. Bush, et al., “Correlation of Excess Power and Helium Production during D₂O and H₂O Electrolysis Using Palladium Cathodes,” Journal of Electroanalytical Chemistry 346 [1993]: 99–117) is single-group and contested on calorimetry and atmospheric-contamination grounds; tier C, reported. ↩
13. The 2004 U.S. DOE review found its reviewers roughly evenly split on whether the excess-heat evidence was convincing and largely unpersuaded that a nuclear process had been demonstrated; the 2019 Google-convened study found no excess heat across 420 samples. U.S. Department of Energy, Report of the Review of Low Energy Nuclear Reactions (December 2004); C. P. Berlinguette et al., “Revisiting the Cold Case of Cold Fusion,” Nature 570 (2019): 45–51. Tier A. ↩
14. Advanced Research Projects Agency–Energy (ARPA-E), “U.S. Department of Energy Announces $10 Million in Funding to Projects Studying Low-Energy Nuclear Reactions,” February 2023. Eight teams, including groups at MIT, Stanford, and Lawrence Berkeley National Laboratory. Institutional primary source; tier A. ↩
15. C. P. Berlinguette et al., “Revisiting the Cold Case of Cold Fusion,” Nature 570 (2019): 45–51. The Google-convened team did not reproduce excess heat (no excess across 420 samples) but identified extreme-loading materials science, and the measurements around it, as underexplored. Peer-reviewed; tier A. ↩