The Forward Physics Facility: TeV Neutrinos from LHC Collisions

The LHC is a proton-proton collider, and its general-purpose detectors (ATLAS and CMS) are designed to measure the particles produced at large angles to the beam line — the central rapidity region where new heavy particles are most likely to be produced and where the experiment’s electromagnetic and hadronic calorimeters give clean signal reconstruction. The very forward direction, at small angles to the beam line, is largely instrumented only for luminosity monitoring and forward-physics studies of cross-sections.

But the LHC also produces large numbers of forward-going neutrinos. Charmed mesons and tau leptons produced in proton-proton collisions decay before being absorbed, and most of the resulting neutrinos are emitted in the forward direction with energies ranging from hundreds of GeV up to several TeV. These neutrinos pass straight through ATLAS and the surrounding rock without interacting, exiting the LHC tunnel essentially unaltered.

The 2023 first detection of collider neutrinos by FASER demonstrated the technique. A small 1-tonne emulsion detector placed 480 metres downstream of the ATLAS interaction point recorded approximately 150 neutrino interactions in its first year of data-taking — the first confirmed observation of neutrinos produced at a collider rather than at a dedicated beam dump. SND@LHC, operating in parallel at the same location, complemented FASER with electronic-detector technology and an enhanced tau-neutrino tagging capability.

The Forward Physics Facility (FPF) is the proposed next step: a dedicated underground cavern about 620 metres downstream of ATLAS that would host a suite of next-generation forward-physics experiments during the HL-LHC era. The FPF would scale the FASER detection volume by orders of magnitude, producing approximately one million collider-made neutrinos per year and opening up a precision-measurement programme that no other facility can match.

This post is about the FPF design, what physics it enables, and how it fits alongside the existing FASER and SND@LHC programmes.

Where the neutrinos come from

The forward neutrino flux at the LHC originates from two principal sources.

Charmed-meson decays are the dominant source of high-energy electron and muon neutrinos. Charmed mesons — $D^0$, $D^{+}$, $D_s$ — are produced in proton-proton collisions through QCD processes involving the gluon and charm quark content of the proton. Their semileptonic and leptonic decays produce neutrinos with energies up to about 4 TeV in the lab frame for HL-LHC operations at 14 TeV centre-of-mass.

The charmed-meson production rate at the LHC is well-known from extensive measurements at ATLAS, CMS, and LHCb, but the forward kinematics are less well-constrained because the central detectors miss most of the forward charm production. FPF measurements of the resulting neutrino spectrum will provide independent constraints on forward charm production that are relevant to the high-energy atmospheric neutrino flux interpretation discussed in our HKKM modelling post.

Tau-lepton production through the chain $D_s\to \tau {\nu }_{\tau }$ followed by tau decay produces the third species of forward neutrino. The tau-neutrino flux at the LHC is the highest-energy source of tau neutrinos available at any laboratory — relevant astrophysical sources at IceCube energies are orders of magnitude higher but with much lower statistics. The FPF tau-neutrino sample provides essentially the only precision measurement of tau-neutrino interactions in the TeV range.

The neutrino flux is highly forward-peaked. At the FPF location 620 metres downstream of the ATLAS interaction point, the relevant angular acceptance is approximately ±1 milliradian — a very narrow cone aligned with the beam line. The expected flux at HL-LHC operations is approximately:

• $10^{12}$ to $10^{13}$ muon neutrinos per year
• $10^{11}$ to $10^{12}$ electron neutrinos per year
• $10^{10}$ to $10^{11}$ tau neutrinos per year

through the FPF cross-section. Detected interaction rates depend on the detector mass; the FLArE detector, planned as a 10-tonne liquid-argon TPC, would record approximately one million interactions per year across all flavours.

The FPF cavern and infrastructure

The Forward Physics Facility is planned as a new underground cavern about 65 metres long and 9 metres in cross-section, located approximately 620 metres downstream of the ATLAS interaction point along the beam line. Construction would require excavating the cavern and a connecting shaft from the surface, with the work scheduled during the HL-LHC shutdown periods to minimise interference with LHC operations.

The cavern site is favourable in several respects: the LHC tunnel between ATLAS and the cavern provides natural shielding against backgrounds from the interaction region; the depth of approximately 90 metres below surface gives sufficient cosmic-ray-muon suppression for the relevant signal-to-background ratios; and the proximity to existing CERN infrastructure simplifies construction and operations.

The total construction cost is estimated at approximately $200 million for the cavern itself, plus the cost of the individual experiments to be hosted in it. Operations would begin in approximately 2030 with first measurements during HL-LHC Run 4.

The FPF experiments

The current FPF design includes several experiments serving different physics goals.

FASERnu2 is the direct successor to the FASERnu emulsion detector that recorded the first collider neutrinos. The next-generation version uses approximately 20 tonnes of emulsion-and-tungsten target, providing exquisite spatial resolution for tracking and identification of tau-neutrino vertices. The emulsion technique is particularly well-suited to tau-neutrino identification through the characteristic kink topology of the prompt-tau decay, which other detectors struggle to see.

FLArE (Forward Liquid Argon Experiment) is a 10-tonne liquid-argon time projection chamber based on DUNE technology, providing electronic readout of neutrino interactions with the full event-topology capability that LArTPCs deliver. FLArE serves both as a high-statistics neutrino-cross-section measurement programme and as a dark-matter search through the elastic-scattering signature of dark matter on argon nuclei.

AdvSND is the next-generation extension of SND@LHC, with improved electronic-detector technology, larger target mass, and enhanced tau-neutrino identification. Operating in conjunction with FASERnu2, it provides complementary measurements with independent systematics.

FORMOSA is a dedicated dark-photon search using scintillator-based detection of the dark-photon decay-to-lepton-pair signature. Its physics target overlaps with SHiP’s parameter space but in a different production mechanism (LHC-produced rather than beam-dump-produced).

MilliQan-FPF and other smaller experiments target specific exotic-particle searches such as millicharged particles, providing additional physics coverage in the FPF cavern.

Physics targets

The FPF physics programme covers several distinct areas.

Precision tau-neutrino measurements are the headline goal for FASERnu2 and AdvSND. Tau-neutrino interactions in the TeV range are essentially unmeasured anywhere else; the LHC’s forward direction is the only source of clean high-statistics tau-neutrino data at these energies. Cross-section measurements provide tests of charged-current universality at high energies and constrain calculations of the tau cross-section that go into IceCube’s astrophysical tau-neutrino flavour-ratio analyses.

Lepton universality at TeV is tested directly by measuring the cross-section ratios for ${\nu }_e$, ${\nu }_{\mu }$, and ${\nu }_{\tau }$ on the same target. The Standard Model predicts these to be identical up to kinematic factors; any deviation would point to flavour-distinguishing new physics in the TeV range — complementary to the τ-decay universality tests at lower energy.

Parton distribution functions at low Bjorken-$x$ are constrained by the forward-charm production rate, which the neutrino flux at FPF directly probes. The relevant parton-distribution-function regime affects predictions for the cosmogenic neutrino flux and for ultra-high-energy cosmic-ray-induced cascade development, so the FPF measurements feed back into astrophysics.

Forward-charm cross-section measurements provide input to atmospheric-neutrino flux calculations at the highest energies, where the prompt component dominated by charmed-meson decays becomes important. The current uncertainty on this prompt flux is a factor of 2-3; FPF measurements would reduce it substantially.

Hidden-sector searches through dark photons, heavy neutral leptons, ALPs, and other long-lived particles produced in LHC collisions. The FPF’s location and detector configuration provide complementary coverage to SHiP, the existing FASER experiments, and the displaced-vertex searches at ATLAS/CMS/LHCb.

How it fits with SHiP

The FPF and SHiP have substantially overlapping but distinct physics targets, particularly in hidden-sector searches.

Production mechanisms differ. SHiP uses beam-dump kinematics, with production dominated by charmed-meson cascades in the dump material. The FPF uses LHC collision kinematics, with production through the gluon and charm quark content of the proton at LHC energies. The two access different parts of the production-cross-section parameter space for hidden particles.

Energy ranges differ. SHiP’s relevant hidden-particle energies are a few GeV up to about 200 GeV. The FPF accesses TeV-scale hidden particles, including the highest-mass dark-matter and HNL candidates that exit the LHC’s central detectors.

Lifetime sensitivity differs. SHiP’s 50-metre decay volume probes lifetimes from microseconds to milliseconds; the FPF’s geometry probes shorter lifetimes (nanoseconds to microseconds) consistent with the higher boost factors at LHC energies.

The two facilities are complementary, with SHiP optimised for the lower-mass, longer-lifetime, smaller-coupling region of parameter space and the FPF optimised for the higher-mass, shorter-lifetime, broader-coupling region.

Where this leads

The Forward Physics Facility represents a strategic recognition that the LHC, beyond being a flagship collider, is also incidentally one of the most powerful sources of high-energy neutrinos. Exploiting this capability with dedicated experiments expands the LHC’s physics output substantially at modest incremental cost relative to the cavern construction.

Beyond the immediate FPF programme, the same forward-collider neutrino technique will extend to the next generation of colliders. The proposed Future Circular Collider (FCC-ee and FCC-hh) at CERN would have its own forward-physics programme by default. The Electron-Ion Collider at Brookhaven National Lab will have forward-neutrino sensitivity through electron-proton scattering, providing complementary information at different beam energies and partons.

The combination of dedicated beam-dump facilities (SHiP), collider-based forward-physics facilities (FPF), and dark-matter direct-detection experiments forms a multi-front search programme for the kinds of weakly-coupled hidden-sector physics that has resisted discovery at central-detector LHC experiments. The TeV-scale forward-physics signal that the LHC produces at no incremental cost is one of the most economical contributions to this broader programme.

Summary

The Forward Physics Facility is a proposed underground cavern about 620 metres downstream of the ATLAS interaction point at the LHC, designed to host a suite of dedicated experiments accessing the very forward direction of LHC proton-proton collisions during the HL-LHC era. The current design includes FASERnu2 (emulsion-based), FLArE (liquid-argon TPC), AdvSND (electronic-detector), and several smaller exotic-particle search experiments, with combined target masses of 20-40 tonnes and expected detection of approximately one million neutrino interactions per year covering energies up to 4 TeV. The physics programme spans precision tau-neutrino cross-section measurements, TeV-scale lepton-universality tests, constraints on forward-charm parton-distribution functions relevant to the atmospheric-neutrino flux, and searches for hidden-sector particles complementary to dedicated facilities like SHiP. Construction would take place during HL-LHC shutdown periods at an estimated cost of $200 million for the cavern, with first physics operations planned for the 2030s. The FPF turns the LHC into an inadvertent neutrino factory at the TeV energy scale, accessing a physics regime that no other facility can match and providing a relatively low-cost extension of the LHC’s overall scientific output.