Protvino to Super-ORCA (P2SO) Overview

• P2SO is a staged long-baseline neutrino experiment combining an upgraded 450 kW beam with a 10× denser Super-ORCA detector for precision oscillation studies.
• It features a 2595 km baseline with multi-GeV matter-enhanced propagation, matching the first oscillation maximum near 5 GeV.
• The setup enables precision measurements of CP violation, mass ordering, and searches for new physics such as non-standard interactions and sterile neutrinos.

Protvino to Super-ORCA (P2SO) is the second-stage long-baseline accelerator-neutrino concept in which a beam from the U-70 synchrotron at Protvino, Russia, is directed to a densified version of the KM3NeT/ORCA detector in the Mediterranean Sea. In the staged P2O/P2SO program, the initial P2O configuration uses ORCA, whereas P2SO couples a 450 kW beam to Super-ORCA, a $\sim 10$ times more-densely instrumented detector optimized for precision measurements of leptonic CP violation over a baseline of about 2595 km, where Earth-matter effects are large and the first oscillation maximum lies near 5 GeV (Akindinov et al., 2019).

1. Origin, staging, and geographical configuration

P2SO emerged as the second stage of the broader Protvino-to-KM3NeT program. The "Letter of Interest for a Neutrino Beam from Protvino to KM3NeT/ORCA" defines a first-stage P2O configuration with ORCA and a 50–90 kW beam, followed by a second-stage P2SO configuration in which the accelerator complex is upgraded to 450 kW and the far detector is replaced by Super-ORCA (Akindinov et al., 2019). In this staging, P2SO is not merely an exposure increase over P2O; it is a combined beam-and-detector upgrade intended specifically for precision CP-phase measurements.

The source is the U-70 proton synchrotron at Protvino, a 70 GeV machine with a circumference of 1.5 km. The far site is the KM3NeT-Fr location in the Mediterranean Sea, about 40 km offshore from Toulon, France. The baseline is generally quoted as $L=2595$ km in later P2SO phenomenology papers, while an early ORCA-focused study used $L=2590$ km for sensitivity calculations to ORCA (Zaborov, 2018). The LoI further specifies that the beam axis is $11.7^\circ$ below the horizon, that the trajectory reaches a deepest point of about 135 km below sea level, and that most of the path lies in the upper mantle, with crustal segments near the source and detector (Akindinov et al., 2019).

The beamline requires new civil infrastructure. The LoI describes an inclined tunnel, extraction in the main hall, a target hall about 30 m deep, horn focusing with reversible polarity, an approximately 180 m decay pipe, an absorber at about 63 m depth, and a near-detector hall at about 90 m depth. The same document places the near detector at about 320 m from the target, or about 120 m downstream of the dump, and identifies it as essential for monitoring flux, energy spectrum, flavor composition, and wrong-sign contamination (Akindinov et al., 2019).

2. Beam, detector, and benchmark operating assumptions

In the design studies, Super-ORCA is a 10$\times$ denser version of ORCA with a 4 Mt fiducial mass, a threshold of about 0.5 GeV, about 100 detected photons per GeV for electromagnetic showers, energy resolution of about 20% for $E_\nu>1$ GeV, and 95%-pure electron-like and muon-like samples (Akindinov et al., 2019). A more detailed detector-performance study reports 115,000 3-inch PMTs per Mton, an energy threshold of about 0.5 GeV for $\nu_e$ CC events and about 0.7 GeV for $\nu_\mu$ CC events, energy resolution better than about 20–25% above 1 GeV, and flavor classification above 1 GeV in which about 95% of $\nu_\mu$ CC events are muon-like while fewer than 5% of $\nu_e$ CC events are classified as muon-like (Hofestädt et al., 2019).

Later phenomenology papers usually do not restate all detector micro-parameters. Instead, they adopt a GLoBES implementation of the P2SO configuration and use a common benchmark: baseline 2595 km, beam power 450 kW, $L=2595$0 protons on target per year, six years of running split as $L=2595$1 years in neutrino and antineutrino modes, and an analysis window of 0.2–10 GeV, with both $L=2595$2 appearance and $L=2595$3 disappearance channels included for neutrinos and antineutrinos (Mishra et al., 2024).

Aspect	Design studies	Common phenomenology benchmark
Baseline	About 2595 km	2595 km
Beam power	Up to 450 kW	450 kW
Exposure convention	Often 10 years, 50% $L=2595$4 / 50% $L=2595$5	6 years, $L=2595$6
Detector	Super-ORCA, 4 Mt fiducial, 10$L=2595$7 denser than ORCA	Same Super-ORCA concept encoded in GLoBES
Energy domain	Threshold around 0.5 GeV; CP-sensitive few-GeV regime	0.2–10 GeV simulation window

The original ORCA-oriented P2O study also quantified beam-driven event statistics. It states that a 90 kW beam would produce about 3000 neutrino events in ORCA every year, and that 3 years at 450 kW correspond to the same exposure factor as 15 years at 90 kW. For ORCA, the same study quotes 5–10$L=2595$8 mass-ordering sensitivity after 1 year at 450 kW, or 5 years at 90 kW, and 2–3$L=2595$9 CP-violation sensitivity after 3 years at 450 kW, or 15 years at 90 kW, with better than $L=2590$0 accuracy on $L=2590$1 (Zaborov, 2018). These numbers apply to ORCA rather than Super-ORCA, but they established the exposure logic subsequently inherited by P2SO studies.

3. Oscillation regime and analysis formalism

The central kinematic feature of P2SO is that the first oscillation maximum for a $L=2590$2 km baseline lies at about 5 GeV. An early P2O study states explicitly that for $L=2590$3 km the first maximum is at $L=2590$4 GeV, while the matter resonance energy in the Earth crust is about 4 GeV (Zaborov, 2018). The LoI emphasizes the same 3–7 GeV window as the region where the broad MSW enhancement and the long baseline combine to generate strong sensitivity to mass ordering and to the interference structure relevant for $L=2590$5 (Akindinov et al., 2019).

A standard approximate expression for the appearance probability used in the P2O/Super-ORCA literature is

$L=2590$6

with

$L=2590$7

This form is used to describe the interplay of intrinsic CP interference and matter enhancement in the few-GeV region (Zaborov, 2018).

In later P2SO simulations, matter is usually treated as constant along the baseline. Several studies adopt $L=2590$8 for P2SO and use the standard matter term $L=2590$9 in the flavor-basis Hamiltonian (Majhi et al., 2022). The long-range-force study writes the propagation Hamiltonian as

$11.7^\circ$0

with $11.7^\circ$1, and then diagonalizes the full Hamiltonian to obtain effective matter-dependent mixing parameters and mass splittings (Mishra et al., 2024). This suggests a common phenomenological architecture in which P2SO new-physics studies probe deformations of the same long-baseline matter Hamiltonian.

Statistically, the P2SO literature is highly standardized. Multiple analyses use GLoBES with modified probability engines and a Poisson log-likelihood of the form

$11.7^\circ$2

combined with pull terms for systematics (Mishra et al., 2024). A frequently used P2SO systematic configuration assigns 5% signal normalization uncertainty in both $11.7^\circ$3 and $11.7^\circ$4 channels, 12% background normalization uncertainty, 11% signal-shape uncertainty, and 4–11% background-shape uncertainty (Majhi et al., 2022).

4. Standard oscillation program

The original Super-ORCA beam study treats P2SO as a precision $11.7^\circ$5 machine. For a 10-year run at 450 kW with equal neutrino and antineutrino exposure, it reports a $11.7^\circ$6 resolution on $11.7^\circ$7 of about $11.7^\circ$8 for $11.7^\circ$9 or $\times$0 and about $\times$1 for $\times$2 or $\times$3 (Hofestädt et al., 2019). The same study shows that after 3 years of neutrino-mode running at 450 kW, the $\times$4 CC sample varies between 8260 events for $\times$5 and 11460 events for $\times$6, and it identifies 50% $\times$7 / 50% $\times$8 running as optimal for resolving the $\times$9–$E_\nu>1$0–$E_\nu>1$1 degeneracy (Hofestädt et al., 2019).

The LoI presents the staged physics case more broadly. For the ORCA stage, it states that P2O can reach at least 3$E_\nu>1$2 mass-ordering sensitivity after about 3 years at 90 kW for any $E_\nu>1$3 and $E_\nu>1$4, and 4–8$E_\nu>1$5 after 1 year at 450 kW (Akindinov et al., 2019). For the Super-ORCA stage, it projects up to about 6$E_\nu>1$6 CP-violation sensitivity and a $E_\nu>1$7 resolution of about $E_\nu>1$8–$E_\nu>1$9 after 10 years at 450 kW, competitive with other planned long-baseline programs (Akindinov et al., 2019).

The division of labor between P2SO and other facilities becomes sharper in later comparative analyses. In the long-range-force study, P2SO is found to outperform T2HKK for $\nu_e$0 octant and mass-ordering determination, while T2HKK remains better for CP-violation sensitivity in both the standard-interaction and long-range-force cases (Mishra et al., 2024). The same analysis identifies a robust feature of direct relevance even in the absence of new physics: the precision on $\nu_e$1 is essentially unchanged in the presence of diagonal long-range-force potentials, with the $\nu_e$2 contours in the $\nu_e$3–$\nu_e$4 plane remaining virtually the same (Mishra et al., 2024). A plausible implication is that the long-baseline spectral structure driving the atmospheric splitting measurement is unusually stable at the P2SO operating point.

5. Beyond-standard-model reach

P2SO has become a reference benchmark for testing nonstandard propagation effects precisely because its baseline amplifies matter-driven distortions while retaining high event statistics near the first maximum.

Long-range flavor forces: In anomaly-free $\nu_e$5 scenarios, the P2SO long-range-force analysis reports the strongest bounds among the compared experiments. At 90% C.L. it finds $\nu_e$6, $\nu_e$7, and $\nu_e$8. For ultralight mediators with $\nu_e$9, these translate into $\nu_\mu$0, $\nu_\mu$1, and $\nu_\mu$2 at 2$\nu_\mu$3 (Mishra et al., 2024).

NSI–LIV discrimination: Because non-standard interactions in matter and CPT-odd Lorentz-invariance-violating coefficients both add energy-independent Hermitian terms to the effective Hamiltonian at fixed density, they are formally equivalent in a single-baseline constant-density setup. The dedicated P2SO study nonetheless finds that discrimination is possible once present and future bounds are imposed asymmetrically on the two scenarios. Its best 3$\nu_\mu$4 case is the diagonal LIV coefficient $\nu_\mu$5, with P2SO achieving separation at $\nu_\mu$6 when NSI are constrained by present bounds and $\nu_\mu$7 when NSI are constrained by future bounds; in most channels P2SO outperforms DUNE (Majhi et al., 2022).

Scalar NSI: For diagonal scalar NSI parameters, one P2SO study reports 3$\nu_\mu$8 limits of $\nu_\mu$9, $\nu_\mu$0, and $\nu_\mu$1. It finds that mass-ordering and CP-violation sensitivities are mostly affected by $\nu_\mu$2, octant sensitivity is mostly affected by $\nu_\mu$3 and $\nu_\mu$4, and the precision of $\nu_\mu$5 deteriorates significantly in the presence of $\nu_\mu$6 and $\nu_\mu$7, while the precision of $\nu_\mu$8 is largely unchanged (Singha et al., 2023). For off-diagonal scalar NSI, another P2SO analysis gives weakest 90% C.L. bounds $\nu_\mu$9, $\nu_e$0, and $\nu_e$1, with exact standard-SNSI mimicry points at about $\nu_e$2, $\nu_e$3, and $\nu_e$4. It further identifies CP-sensitivity washout points near $\nu_e$5, $\nu_e$6, and $\nu_e$7 for specific phases (Pusty et al., 2024).

Sterile neutrinos in $\nu_e$8: In the KM3NeT long-baseline options paper, P2SO is the strongest of the P2O/upgraded-P2O/P2SO configurations and is described as either comparable or better than DUNE. The study emphasizes that the near detector is very important, that adding it improves sensitivity relative to a far-detector-only analysis, and that P2SO is most competitive around $\nu_e$9 because the near-detector flux overlap is favorable in that region (Singha et al., 2022).

Decoherence: The decoherence study examines two Lindblad formalisms. For P2SO, the 3$L=2595$00 bounds are $L=2595$01 in matter for Formalism-A and $L=2595$02 in matter for Formalism-B. The same paper concludes that P2SO is more sensitive than DUNE in Formalism-A with matter, while DUNE is stronger in Formalism-B (Majhi et al., 22 Apr 2026).

Non-unitarity: In a triangular non-unitary PMNS parameterization, P2SO is found to be stronger than DUNE for $L=2595$03 and especially $L=2595$04, whereas DUNE is stronger for $L=2595$05 and $L=2595$06. At 90% C.L. in the six-degree-of-freedom fit, the paper reports lower bounds $L=2595$07 and $L=2595$08 for P2SO, and explicitly notes that P2SO improves the current bound on $L=2595$09 (Pusty et al., 1 Mar 2026).

Large extra dimensions and neutrino decay: For LED, P2SO alone gives $L=2595$10 at 90% C.L. in the weakest quoted scenario with all parameters free and systematics included, while the combined DUNE+T2HK+P2SO fit yields $L=2595$11 in the same scenario. The same study finds that LED has minimal impact on CP violation, mass ordering, and octant at realistic small $L=2595$12 (Panda et al., 2024). For invisible decay of $L=2595$13, P2SO gives $L=2595$14 at 3$L=2595$15, better than MOMENT and ESSnuSB but weaker than T2HK and DUNE; it also shows that decay can degrade CP sensitivity while altering octant sensitivity in a non-trivial manner (Panda et al., 2024).

6. Comparative position, complementarity, and recurrent caveats

Across the literature, P2SO is usually presented as complementary rather than redundant relative to DUNE, T2HK, and T2HKK. The underlying reason is stable across very different theoretical scenarios: P2SO combines a 2595 km baseline, strong matter effects, and an operating point near the first oscillation maximum, which enhances sensitivity to flavor-diagonal matter-like perturbations, disappearance-channel distortions, and ordering-dependent effects (Akindinov et al., 2019). By contrast, experiments with shorter baselines or stronger second-maximum leverage often retain an advantage for direct CP-violation sensitivity.

This complementarity appears repeatedly in head-to-head comparisons. P2SO is stronger than T2HKK for long-range-force bounds, octant, and ordering, while T2HKK is better for CP violation (Mishra et al., 2024). P2SO is competitive with or better than DUNE in sterile-neutrino studies around $L=2595$16, but DUNE is favored around $L=2595$17 because of the near-detector flux alignment (Singha et al., 2022). In non-unitarity studies, P2SO dominates the matter-driven $L=2595$18 sector, whereas DUNE dominates the appearance-driven $L=2595$19 and $L=2595$20 sectors (Pusty et al., 1 Mar 2026).

Several caveats also recur. Many phenomenological studies use constant matter density, typically with $L=2595$21, and rely on a GLoBES implementation rather than a fully rederived detector-response model (Majhi et al., 2022). Many papers state explicitly that detailed detector resolutions and binning are inherited from earlier configurations rather than itemized anew (Mishra et al., 2024). Near-detector information is repeatedly identified as essential for flux, cross-section, wrong-sign, and sterile-oscillation control (Akindinov et al., 2019). This suggests that the eventual realized sensitivity of P2SO will depend not only on baseline and far-detector performance, but also on the maturity of the beamline, the systematic-control program, and the extent to which timing, directionality, and spectral-shape information are incorporated into the final analysis chains.

Within that framework, P2SO occupies a distinctive place in future long-baseline neutrino physics. Its defining attributes are not simply high statistics or large fiducial mass, but the conjunction of a very long baseline, multi-GeV matter-enhanced propagation, and a Super-ORCA detector designed to extend underwater Cherenkov reconstruction into the sub-few-GeV regime. That combination makes it simultaneously a precision oscillation facility and a stringent probe of matter-sensitive departures from three-flavor neutrino propagation.