In 2011, a French team led by Thierry Mention published a paper that quietly upended one corner of neutrino physics. The paper presented a new calculation of the electron antineutrino flux produced by nuclear reactors — built up from the latest measurements of the electron beta-decay spectra of the dominant fission products in reactor fuel — and found that the new predicted flux was about 6% higher than the value used in earlier analyses.
The implication was striking. The world’s short-baseline reactor antineutrino experiments — which had been running for decades — had been measuring rates that, in retrospect, were about 6% below the new theoretical predictions. The systematic deficit was consistent across many experiments at different reactors using different detectors. This became known as the reactor antineutrino anomaly, or RAA.
A second puzzle emerged a few years later. Several precision experiments — Daya Bay, RENO, Double Chooz — observed an unexpected excess of events near 5 MeV in the antineutrino energy spectrum, where the measured rate is roughly 10% above predictions. The location of the excess was nearly identical across the three experiments, strongly suggesting it was a real feature of reactor emission rather than detector artifacts.
The two puzzles together — the 6% deficit and the 5 MeV bump — have been the subject of nearly fifteen years of intense experimental and theoretical follow-up. The picture has narrowed substantially. It is not yet fully resolved.
What reactor antineutrinos are
Nuclear reactors produce electron antineutrinos as byproducts of nuclear fission. The four dominant fissioning isotopes in a typical light-water reactor are uranium-235, uranium-238, plutonium-239, and plutonium-241. Each fission produces, on average, about 200 MeV of energy and roughly six antineutrinos as the resulting unstable fragments beta-decay toward stability.
A typical commercial reactor emits roughly 6 × 10²⁰ antineutrinos per second with energies up to about 10 MeV. Most of these are produced by neutron-rich fission fragments cascading through chains of beta decays. The total emitted spectrum is the weighted sum of contributions from thousands of individual fission-product beta decays.
Predicting that spectrum precisely is hard. The two main approaches are:
The summation method. Calculate the spectrum from first principles by summing over every individual beta-decay branch of every fission fragment. The data needed is the branching fractions and end-point energies of thousands of decay branches.
The conversion method. Measure the total electron spectrum from a fissioning sample in the laboratory and convert it to an antineutrino spectrum using the kinematics of beta decay. This method, used by Schreckenbach and colleagues in the 1980s, was the basis of the predictions Mention’s 2011 paper revised.
The conversion method was widely considered the more accurate approach for years, partly because it depended on direct measurements rather than incomplete nuclear data tables. But the agreement between the two methods is imperfect, and the recent re-evaluations of the conversion approach are what produced the 6% upward shift that defined the anomaly.
The sterile-neutrino hypothesis
The most exciting interpretation of the reactor antineutrino anomaly was that short-baseline oscillation into a sterile neutrino was depleting the detected flux. If a sterile neutrino species with a mass-squared difference of roughly 1 eV² exists and mixes weakly with the electron neutrino, then reactor antineutrinos with energies of a few MeV would partially oscillate into the sterile state over baselines of about 10 meters — exactly the kind of short distance that older reactor experiments could not resolve.
This connected the reactor anomaly to two other anomalies: the LSND and MiniBooNE results from accelerators, and the gallium anomaly from solar neutrino calibration experiments. All three pointed loosely toward a sterile neutrino at around 1 eV.
If the sterile-neutrino interpretation was correct, it would imply a fourth neutrino species, new physics beyond the Standard Model, and significant implications for cosmology. It would also be straightforward to test: build a reactor experiment with a detector that moves through different baselines and measure whether the antineutrino spectrum oscillates with distance.
What very-short-baseline experiments found
Several experiments were built specifically to test the sterile-neutrino hypothesis with reactor antineutrinos at very short baselines.
PROSPECT at the Oak Ridge National Laboratory’s High Flux Isotope Reactor used a segmented liquid scintillator detector at distances of 7–9 meters from a compact research reactor.
STEREO at the Institut Laue-Langevin in Grenoble, France, used a similar concept at distances of 9–11 meters.
DANSS at the Kalinin Nuclear Power Plant in Russia used a segmented plastic scintillator detector that could be moved through three different distances (10.7, 11.7, 12.7 meters) from the reactor core.
Neutrino-4 at the SM-3 reactor in Dimitrovgrad, Russia, claimed a positive sterile-neutrino signal — but this claim has been disputed and is generally not considered confirmed.
Across the broader experimental program, the data are consistent with no short-baseline oscillation at the parameters needed to explain the reactor anomaly via sterile neutrinos. The simplest sterile-neutrino interpretation is now disfavored, with combined limits excluding most of the parameter space that would have explained the original deficit.
Where the deficit is going
If the deficit is not sterile-neutrino oscillation, it must be a problem with the antineutrino flux prediction.
Recent work suggests that the conversion-method spectrum is biased, principally because the underlying electron spectra of the Schreckenbach reference measurements — which were made in the 1980s — may have residual systematic errors. New summation calculations, using updated nuclear-database information, predict a flux that is closer to the measured values, removing much of the deficit.
A separate strand of evidence comes from comparing the antineutrino spectrum from reactors with different fuel compositions. Some experiments have measured a “fuel-dependent” deficit, with the uranium-235 contribution apparently lower than expected while plutonium-239 matches predictions better. This pattern is hard to explain with a generic oscillation hypothesis but fits naturally with errors specific to the uranium-235 prediction in the conversion-method calculations.
The current consensus is that most of the 6% deficit is a flux-prediction issue, not new physics. A residual disagreement of a few percent remains, and the field continues to refine both the predictions and the measurements.
What the 5 MeV bump might be
The 5 MeV bump is a different puzzle. It appears in the energy region where the predicted spectrum has a particular shape sensitive to the contributions of specific fission fragments with high-energy beta-decay endpoints. The experiments see a real excess of events at this energy, consistent across multiple reactors.
The leading explanation now is that the original spectrum calculations did not adequately model forbidden beta-decay transitions. In a forbidden transition, the beta decay proceeds at a rate suppressed by selection rules, and the spectrum shape differs from the simple allowed-transition shape used in most calculations. Several fission products in the 5 MeV region undergo forbidden transitions; if their spectra were modeled with the simpler allowed form, the predicted total spectrum would be biased downward exactly where the data shows the bump.
Updated nuclear-data calculations now incorporating proper forbidden-transition treatment do appear to predict the 5 MeV bump much better. The puzzle is, increasingly, looking like a triumph of careful nuclear data rather than evidence for new physics.
A useful anomaly
Even if both puzzles ultimately resolve as nuclear-physics issues rather than new physics, the reactor antineutrino anomaly has been an enormously productive period for neutrino physics. The 2011 paper triggered:
- A generation of new short-baseline reactor experiments (PROSPECT, STEREO, DANSS, Neutrino-4, NEOS, NUCIFER).
- Major theoretical work on reactor antineutrino spectra and nuclear-data evaluations.
- Tighter constraints on the sterile-neutrino parameter space at 1 eV.
- A precision benchmark that future reactor experiments — including JUNO — can build on.
The anomaly itself may turn out not to be new physics. The collateral benefit — a much sharper picture of how reactors actually emit antineutrinos, and tighter sterile-neutrino limits — has been substantial.
Why this kind of puzzle matters
Most “anomalies” in particle physics either get confirmed as discoveries within a few years or evaporate within a decade. The reactor antineutrino anomaly has been unusual in lasting fifteen years and producing a slow, careful narrowing of the parameter space rather than a clean resolution.
That kind of patient, multi-experiment investigation is sometimes underappreciated. It rarely produces a Nobel-worthy “discovery.” But it sharpens the tools — both experimental and theoretical — that the field uses for everything else. When JUNO begins precision measurements of reactor antineutrinos in the late 2020s, it will inherit a much better understanding of the reactor antineutrino spectrum than it would have had without fifteen years of anomaly-driven follow-up. The investment will pay off in precision wherever the answer eventually settles.
For the closely-related sterile-neutrino story, see Sterile neutrinos: a stubborn maybe and MicroBooNE. For the original θ₁₃-measuring reactor experiments, see Daya Bay and RENO. For the next-generation reactor experiment built around mass ordering rather than anomalies, see JUNO.
Frequently asked
What is the reactor antineutrino anomaly?
Beginning in 2011, a recalculation of the antineutrino flux expected from nuclear reactors found values about 6% higher than previously thought. When compared to actual measurements from short-baseline reactor experiments, the data showed a corresponding 6% deficit — fewer antineutrinos arriving than calculations predicted. The deficit has remained stubbornly consistent across multiple experiments and reactor types.
What is the 5 MeV bump?
Several reactor experiments — Daya Bay, RENO, and Double Chooz — observed an unexpected excess of antineutrino events around 5 MeV, where the measured spectrum is roughly 10% higher than predictions. The location and shape of the bump are similar across experiments, suggesting it is a real feature of reactor antineutrino emission rather than an experimental artifact.
Could a sterile neutrino explain the reactor anomaly?
Yes, in principle. A sterile neutrino with mass around 1 eV would cause some reactor electron antineutrinos to oscillate into a non-detectable flavor at short baselines, producing exactly the kind of deficit seen. However, very-short-baseline experiments designed to directly test this idea — PROSPECT, STEREO, and others — have ruled out the simplest sterile-neutrino interpretation, leaving alternative explanations on the table.
What are the alternative explanations?
Most current evidence points to errors in the predicted antineutrino spectrum rather than oscillation. The original flux calculations relied on summed electron-spectrum measurements from fissioning isotopes, and recent re-analyses suggest those measurements may have systematic biases. A separate re-analysis using a different method ('summation') largely removes the deficit. The 5 MeV bump appears to be a feature of forbidden beta-decay transitions in fission fragments that the original calculations did not adequately model.
Is the reactor antineutrino anomaly resolved?
Not fully. The picture has been narrowed significantly: the simplest sterile-neutrino interpretation is largely excluded, and most of the deficit can be attributed to flux-calculation issues. But residual disagreement between data and the best current predictions remains at the few-percent level, and the precise origin of the 5 MeV bump is still being modeled.
Cite this article 5 formats
APA
Neutrino Times Editorial Team. (2025, October 27). The reactor antineutrino anomaly: a 6% deficit that won't go away. Neutrino Times. https://neutrino-times.com/articles/reactor-antineutrino-anomaly-5-mev-bump/
Chicago
Neutrino Times Editorial Team. "The reactor antineutrino anomaly: a 6% deficit that won't go away." Neutrino Times, October 27, 2025. https://neutrino-times.com/articles/reactor-antineutrino-anomaly-5-mev-bump/.
MLA
Neutrino Times Editorial Team. "The reactor antineutrino anomaly: a 6% deficit that won't go away." Neutrino Times, 27 Oct. 2025, https://neutrino-times.com/articles/reactor-antineutrino-anomaly-5-mev-bump/.
BibTeX
@misc{neutrino-times-reactor-antineutrino-anomaly-5-mev-bump,
author = {Neutrino Times Editorial Team},
title = {The reactor antineutrino anomaly: a 6% deficit that won't go away},
howpublished = {Neutrino Times},
year = {2025},
month = {oct},
url = {https://neutrino-times.com/articles/reactor-antineutrino-anomaly-5-mev-bump/},
note = {Accessed: 2025-10-27}
} RIS
TY - GEN TI - The reactor antineutrino anomaly: a 6% deficit that won't go away AU - Neutrino Times Editorial Team PY - 2025 DA - 2025-10-27 PB - Neutrino Times UR - https://neutrino-times.com/articles/reactor-antineutrino-anomaly-5-mev-bump/ ER -