The calibration house in JUNO

A. Freegard; A. Takenaka; H. Lai; J. Huang; J. Hui; J. Liu; P. Zhang; R. Li; T. Zhang; Y. Liao

REVIEW 3 major objections 6 minor 19 references

The calibration house in JUNO

T0 review · 3 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper establishes that the JUNO calibration house, a sealed stainless-steel chamber connecting calibration equipment to the central detector, maintains oxygen below 10 ppm and radon below 15 mBq/m³, with a total leak rate below…

desk verdict A solid technical report on the JUNO calibration house; the radon budget has a units bug and a thin safety margin, but the measurements and design work are real. read the letter →

arxiv 2507.09208 v1 pith:FJROJ5TH submitted 2025-07-12 physics.ins-det hep-ex

J. Hui , R. Li , Y. Wu , T. Zhang , Z. Chen , A. Freegard , J. Huang , H. Lai

show 7 more authors

Y. Liao J. Liu Y. Meng A. Takenaka Z. Xiang P. Zhang Y. Zhang

This is my paper · ORCID

classification physics.ins-dethep-ex

keywords calibrationhouseJUNOradonconcentrationcontroloxygennitrogenflushleakrateliquidscintillatorbackgroundsuppression

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper reports on the calibration house, a sealed stainless-steel chamber that connects calibration equipment to the central detector of the JUNO neutrino experiment. The paper shows that the house meets its design requirements: oxygen concentration can be kept below 10 ppm, radon concentration below 15 mBq/m³, and total leak rate below 2.9×10⁻⁵ mbar·L/s. Onsite installation confirmed that the mechanical and electrical interfaces work with the calibration sub-systems. The authors also present two radon diffusion models that convert the measured house radon concentration into an expected background contribution to the detector, concluding the contribution is within the 0.2 mBq/m³ budget.

What carries the argument

The load-bearing objects are the nitrogen flush system and the radon diffusion models. The flush system, with a specific inlet/outlet layout (layout 5) that avoids flow shortcuts, reduces oxygen from 20% to below 5 ppm within about an hour. The radon diffusion model 2 uses the complementary error function, $R = \operatorname{erfc}(h/\sqrt{4a\tau})$, with $h = 4$ m of liquid scintillator in the chimney, $a = 1.49\times10^{-9}$ m²/s, and $\tau = 20$ years, giving $R \approx 0.35\%$ as the fraction of house radon that reaches the detector. This factor, applied to the measured radon concentration, translates the house level into a detector background.

What would settle it

Measure the radon activity in the liquid scintillator near the bottom of the chimney during steady-state operation and compare it with the radon concentration in the calibration house; if the ratio exceeds the 0.35% prediction by more than the measurement errors, the diffusion-only model is wrong and the house requirement would need to be tightened.

Watch

Extended reading notes

Core claim

The central claim is that the calibration house, as built, meets JUNO's requirements for background and leak tightness. The evidence includes nitrogen flush tests showing oxygen falls below 5 ppm, radon measurements with a customized detector showing concentrations as low as 8±4 mBq/m³ with nitrogen flush, a glove box that passed all operational tests, and an onsite leak check of 21 flanges plus feedthroughs giving a total leak rate below 2.9×10⁻⁵ mbar·L/s. The paper further argues, using a diffusive transport model through 4 m of quiescent liquid scintillator in the chimney, that the radon entering the central detector is about 0.35% of the house concentration, so the measured house level corresponds to 0.08±0.04 mBq/m³, below the 0.2 mBq/m³ budget.

Load-bearing premise

The radon entering the detector is assumed to travel only by diffusion through 4 meters of still liquid scintillator in the chimney, with no convection and with radon and its decay products diffusing at the same rate; if the liquid moves or mixes, the true radon entry could be higher than the 0.35% factor used.

Editorial extensions

If this is right

JUNO can deploy the four calibration sub-systems without compromising the liquid scintillator's oxygen and radon budgets.
The double O-ring nitrogen pressure-drop leak check method is suitable for large thin-wall chambers that cannot be vacuum-tested.
The radon contribution from the calibration house is small: 0.08±0.04 mBq/m³, about 40% of the 0.2 mBq/m³ budget.
Cleanliness of the chamber matters: dust contributed 22±15 mBq/m³ before cleaning, comparable to the contribution from the calibration sub-systems themselves.
Glove box operations can be performed without detectable oxygen ingress during typical 30-minute interventions.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The 0.35% diffusion factor is the largest leverage on the background budget; if even mild convection exists in the chimney or during detector filling, the factor could be closer to the model-1 estimate of 1%, still within the budget but with a thinner margin.
The nitrogen-flush layout-selection method, using multiple oxygen sensors to find a layout that avoids flow shortcuts, could be applied to other large detector volumes with complex internal geometry.
The measurement that dust adds 22±15 mBq/m³ suggests that maintaining cleanliness after installation is as important as selecting low-radioactivity materials; periodic re-cleaning or radon monitoring might be warranted.
If the radon diffusion coefficient in liquid scintillator were measured under flow or temperature-gradient conditions, the 20-year integration time assumption could be replaced with a more realistic operating scenario.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague,

The JUNO calibration house paper is a solid, straightforward technical report. What's new is the first detailed description of the house itself: the interface design, the five nitrogen flush layouts compared empirically, and the onsite oxygen/radon/leak measurements. The flush layout comparison using oxygen as a tracer is a nice piece of practical engineering, and the leak check strategy (SF6 sniffing for feedthroughs, double O-ring pressure drop for flanges) is sensible. The measured oxygen <5 ppm and radon [2,15] mBq/m3 support the headline claims.

The soft spot is the radon budget arithmetic in Section 4.2.3. The paper compares 0.08 mBq/m3 (8 mBq/m3 times 1%) against a 'requirement' of 0.2 mBq/m3, but that requirement was derived by dividing the 0.005 Bq detector budget by the 25 m3 house volume—which is only valid if all radon in the house enters the CD. The correct check is C_house * epsilon * V_house < 0.005 Bq. With C = 8 mBq/m3, epsilon = 1%, V = 25 m3, you get 0.002 Bq, which passes but only by a factor of 2.5, not by the large margin the presentation implies. The paper's own arithmetic shows the same factor of 2.5, so the stress-test's claim of a hidden factor of 100 is overstated, but the dimensional muddle should be fixed in revision. More importantly, the margin is thin enough that the no-convection assumption becomes load-bearing. If mixing or convection in the chimney raises the effective transfer fraction above ~2.5%, the 0.005 Bq budget is broken. The authors do state the assumption explicitly, and both models (1% geometric, 0.35% diffusive) are presented as simple estimates, so I don't think this is fatal—but a sensitivity study around epsilon would strengthen the claim considerably.

Overall: the paper is honest, clearly written, and the measurements look real. For a JINST-style technical note, it's a solid accept after minor revision. I'd want the radon budget section rewritten to get the units straight and to acknowledge the margin is only a factor of ~2.5. The calibration house itself appears to work as designed.

Recommendation: send to peer review; accept with minor revisions.

Referee Report

3 major / 6 minor

Summary. The paper describes the design, installation, and commissioning tests of the calibration house for the JUNO central detector, focusing on mechanical interfaces, glove boxes, cable routing, a nitrogen flush system, and control of oxygen and radon concentrations. The authors report that the oxygen concentration can be maintained below 5 ppm, the radon concentration in the flushed configuration is [2,15] mBq/m3 at 90% C.L., and the total onsite leak rate is below 2.9e-5 mbar L/s, meeting the design requirement of 4e-5 mbar L/s. To connect the measured house radon concentration to the detector background budget, the paper introduces two diffusion models that yield conversion factors of about 1% and 0.35%, and then compares the measured concentration times the 1% factor (0.08±0.04 mBq/m3) with a derived requirement of 0.2 mBq/m3.

Significance. The manuscript is a useful engineering and commissioning report for an auxiliary system of JUNO. Its strengths are the direct measurements: oxygen below 5 ppm with a calibrated sensor, radon measured with a dedicated detector and reported with a proper confidence interval, and leak checks performed with calibrated instruments meeting the stated limits. The glove-box functionality tests and the cable-distribution validation are also concrete and reproducible. If the radon-budget analysis is clarified and the sensitivity of the conversion factor is properly quantified, the paper will serve as a solid reference for the JUNO collaboration and for similar low-background calibration facilities.

major comments (3)

[Section 2 and Section 4.2] The radon concentration requirement is stated as 20 mBq/m3 in Section 2, but Section 4.2 derives a requirement of 0.2 mBq/m3 (0.005 Bq divided by the 25 m3 house volume). These two numbers differ by a factor of 100, and the manuscript does not explain which is the actual design requirement or how they are related. If the 0.2 mBq/m3 value is intended as the requirement on the effective concentration after the transfer fraction, the text must say so explicitly; otherwise the reader cannot tell whether the measured 8±4 mBq/m3 meets the requirement directly or only after applying the conversion model.
[Section 4.2.3] The comparison of 0.08±0.04 mBq/m3 with 0.2 mBq/m3 is dimensionally consistent only if the 0.2 mBq/m3 is understood as the allowable product C_house × ε (with ε the transfer fraction). The resulting margin is a factor of about 2.5, not the factor of 100 that a casual comparison of 8 mBq/m3 with 20 mBq/m3 would suggest. The authors should present this margin explicitly and discuss how the conclusion changes if ε is larger than the nominal 1% value.
[Section 4.2.2] The radon budget rests on the assumption that the 4 m of liquid scintillator in the chimney is quiescent with no convection during data taking and calibration deployments. Calibration operations involve moving cables, source deployment, and possible thermal gradients, all of which could mix the LS and increase the transfer fraction beyond the 1% or 0.35% assumed. The paper should either justify the no-convection assumption for the actual operating conditions or provide a conservative upper bound on ε that accounts for mixing. In addition, the erfc solution in Eq. (4.1) is derived for a stable species and does not explicitly account for the radioactive decay of 222Rn; the authors should clarify how decay is treated and whether the model over- or under-estimates the radon reaching the detector.

minor comments (6)

[Abstract] The abstract states the radon concentration can be kept below 15 mBq/m3; this is the upper end of the 90% C.L. interval [2,15] mBq/m3, not the central value. Consider reporting the central value and interval for precision.
[Section 4.2] The phrase 'equivalent to 0.2 mBq/m3' should state that this is obtained by dividing the 0.005 Bq budget by the internal volume of the calibration house (25 m3), so that the reader understands the implicit assumption behind the equivalence.
[Section 4.2.1] The surface-area model assumes that every radon atom colliding with the LS surface in the chimney is absorbed and enters the CD, while collisions with the house walls lead to reflection. This is a strong assumption about the sticking coefficient; a brief justification or a reference to a similar treatment would strengthen the model.
[Section 4.2.2] The diffusion coefficient from reference [21] is given without context. State its measurement conditions and whether it applies to radon in the JUNO liquid scintillator at the relevant temperature.
[Section 4.2.3] The measured radon concentration after flushing is reported as 8±4 mBq/m3, with a relative uncertainty of 50%. It would be informative to also report the 90% C.L. upper limit (15 mBq/m3) in the budget comparison, since the central value alone may underestimate the risk.
[General] There are a number of typographical and grammatical errors, for example 'large' in Section 3.1 and 'the CLS SS cable jump out' in Table 1. A careful proofreading pass would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the validation claims rest on direct measurements and externally specified benchmarks, not on fitted parameters or self-citation chains.

full rationale

The paper's main claims are validation results, not predictions derived from a fitted model. The oxygen claim (Section 4.1) is a direct measurement after nitrogen flushing, with the leak-rate contribution assessed from a separately measured leak rate. The radon claim (Sections 4.2.1-4.2.3) combines a measured house concentration (8 +/- 4 mBq/m^3 after flushing) with two explicitly stated transfer models (1% surface-absorption model and 0.35% pure-diffusion model). The diffusion coefficient used in model 2 comes from an external measurement (reference [21]) that does not incorporate the target result, and no parameter is fitted to the pass/fail comparison. The 0.2 mBq/m^3 requirement is derived from a stated JUNO radioactivity budget (reference [16]) divided by the house volume; comparing the effective contribution concentration (0.08 mBq/m^3, i.e., 8 mBq/m^3 x 1%) to this requirement is dimensionally equivalent to comparing total activities, since both sides carry the same house volume factor. The leak-rate claim (Section 4.3) is a sum of independently measured flange and feedthrough leak rates. Self-citations appear for the JUNO specifications, the LS radon diffusion measurement, and a radon-detector systematic, but none of these supply the conclusion itself; they are external inputs with stated assumptions. The no-convection assumption in model 2 is a physical assumption that, if wrong, would affect the margin, but that is a correctness risk rather than circularity.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The central claims rest on measured values and a few simple diffusion-model assumptions, not on fitted parameters. No free parameters are introduced; the diffusion coefficient and background requirements come from prior literature.

assumptions (6)

domain assumption No convection occurs in the upper 4 m of liquid scintillator within the chimney during stable data acquisition.
Section 4.2.2. Used to justify diffusion-only transport of radon from the calibration house to the CD; if convection exists, the conversion factor R=0.35% is too small and the radon budget could be exceeded.
domain assumption Radon and its decay products share the same diffusion coefficient in the liquid scintillator.
Section 4.2.2. Needed for Eq. (4.1) to apply to both 222Rn and its decay products; stated without justification in the paper.
domain assumption Radon moves freely in the calibration house and is absorbed only when it hits the liquid scintillator surface in the chimney; collisions with house walls reflect it back.
Section 4.2.1. This idealization gives the 1% geometric conversion factor; the real absorption and mixing behavior is likely more complex.
domain assumption The diffusion coefficient a=1.49e-9 m^2/s for radon in JUNO liquid scintillator is taken from reference [21].
Section 4.2.2, cited external measurement; its validity for the actual JUNO LS is assumed.
domain assumption The calibration house contributes at most 20% of the total JUNO radon budget.
Section 4.2. This allocation sets the 0.2 mBq/m3 requirement; it is an asserted design choice, not derived.
domain assumption The limits 10^-9 Bq/L for 222Rn and 10^-24 g/g for 210Pb in the liquid scintillator are taken from reference [16].
Section 4.2, basis for the radon budget calculations.

how reviews work

0 comments

Cite this review

Pith. "Pith review of The calibration house in JUNO." pith.science (2026). https://pith.science/paper/FJROJ5TH

@misc{pith2026250709208,
  author       = {Pith},
  title        = {Pith review of: The calibration house in JUNO},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FJROJ5TH}},
  note         = {Machine review of arXiv:2507.09208}
}

abstract

As an auxiliary system within the calibration system of the Jiangmen Underground Neutrino Observatory, a calibration house is designed to provide interfaces for connecting the central detector and accommodating various calibration sub-systems. Onsite installation has demonstrated that the calibration house interfaces are capable of effectively connecting to the central detector and supporting the installation of complex and sophisticated calibration sub-systems. Additionally, controlling the levels of radon and oxygen within the calibration house is critical. Radon can increase the experimental background, while oxygen can degrade the quality of the liquid scintillator. The oxygen concentration can be maintained at levels below 10 parts per million, and the radon concentration can be kept below 15 mBq/m$^{3}$. This paper will provide detailed information on the calibration house and its methods for radon and oxygen concentration control.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

19 extracted references · 17 canonical work pages

[2]

Abusleme, T

A. Abusleme, T. Adam, S. Ahmad, R. Ahmed, S. Aiello, M. Akram et al.,The design and technology development of the juno central detector,The European Physical Journal Plus139(2024) 1128

work page 2024
[3]

L. Zhan, Y. Wang, J. Cao and L. Wen,Experimental Requirements to Determine the Neutrino Mass Hierarchy Using Reactor Neutrinos,Phys. Rev. D79(2009) 073007 [0901.2976]. [4]JUNOcollaboration,Calibration Strategy of the JUNO Experiment,JHEP2021(2021) 4 [2011.06405]

arXiv 2009
[5]

Laser Calibration System in JUNO

Y. Zhang, J. Liu, M. Xiao, F. Zhang and T. Zhang,Laser Calibration System in JUNO,JINST14 (2019) P01009 [1811.00354]

work page Pith review arXiv 2019
[6]

A.Takenakaetal.,Customizedcalibration sourcesinthe JUNOexperiment,JINST19(2024)P12019 [2410.01571]

work page Pith review arXiv 2024
[7]

The Automatic Calibration Unit in JUNO

J. Hui et al.,The automatic calibration unit in JUNO,JINST16(2021) T08008 [2104.02579]

work page Pith review arXiv 2021
[8]

Y. Guo, Q. Zhang, F. Zhang, M. Xiao, J. Liu and E. Qu,Design of the Guide Tube Calibration System for the JUNO experiment,JINST14(2019) T09005 [1905.02077]

work page Pith review arXiv 2019
[9]

Y. Guo, K. Zhu, Q. Zhang, F. Zhang, Y. Meng, J. Liu et al.,Construction and Simulation Bias Study of The Guide Tube Calibration System for JUNO,JINST16(2021) T07005 [2103.04602]

work page Pith review arXiv 2021
[10]

Cable Loop Calibration System for Jiangmen Underground Neutrino Observatory

Y. Zhang, J. Hui, J. Liu, M. Xiao, T. Zhang, F. Zhang et al.,Cable loop calibration system for Jiangmen Underground Neutrino Observatory,Nucl. Instrum. Meth. A988(2021) 164867 [2011.02183]

work page Pith review arXiv 2021

Show all 19 references

[11]

K.Feng, D.Li, Y.Shi, K.QinandK.Luo,Anovelremotelyoperatedvehicleasthecalibrationsystem in JUNO,JINST13(2018) T12001

2018
[12]

Zhu, J.-L

G.-L. Zhu, J.-L. Liu, Q. Wang, M.-J. Xiao and T. Zhang,Ultrasonic positioning system for the calibration of central detector,Nucl. Sci. Tech.30(2019) 5

2019
[13]

Teng, J.-L

D. Teng, J.-L. Liu, G.-L. Zhu, Y. Meng, Y.-y. Zhang, T. Zhang et al.,Low-radioactivity ultrasonic hydrophone used in positioning system for Jiangmen Underground Neutrino Observatory,Nucl. Sci. Tech.33(2022) 76

2022
[14]

G. Zhu, W. Yang, D. Teng, Q. Wang, J. Hui and J. Lian,Experiments of Ultrasonic Positioning System with Symmetrical Array Used in Jiangmen Underground Neutrino Observatory,Symmetry16(2024) 1218

2024
[15]

Liu,private communication, Tech

J. Liu,private communication, Tech. Rep. [16]JUNOcollaboration,Radioactivity control strategy for the JUNO detector,JHEP11(2021) 102 [2107.03669]. [17]JUNOcollaboration,The Top Tracker of the JUNO Experiment,PoSEPS-HEP2023(2024) 194

2021
[18]

Zhao,private communication, Tech

K. Zhao,private communication, Tech. Rep
[19]

Ling et al.,JUNO high purity nitrogen plant,Appl

X. Ling et al.,JUNO high purity nitrogen plant,Appl. Radiat. Isot.208(2024) 111305. – 16 –

2024
[20]

Zhao,New Concept Physics Course: Thermodynamics (Second Edition), Higher Education Press (2005)

K. Zhao,New Concept Physics Course: Thermodynamics (Second Edition), Higher Education Press (2005)

2005
[21]

Z. Xu, C. Guo, J. Liu, Y. Zhang, P. Zhang, C. Yang et al.,Research of radon diffusion behavior in liquid scintillator,JINST18(2023) P04006 [2301.06982]. [22]PandaX-4Tcollaboration,Low radioactive material screening and background control for the PandaX-4T experiment,JHEP06(202...

2023 arXiv
[23]

Y.Wu, L.Si,Z.Qian,Y.Yun,Y.Meng,J.Liuetal.,Developmentofhigh-sensitivityradonemanation measurement systems with surface treatment optimization,arXiv:2503.00739

arXiv
[24]

Feldman and R.D

G.J. Feldman and R.D. Cousins,Unified approach to the classical statistical analysis of small signals, Phys. Rev. D57(1998) 3873. – 17 –

1998

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [2]

Abusleme, T

A. Abusleme, T. Adam, S. Ahmad, R. Ahmed, S. Aiello, M. Akram et al.,The design and technology development of the juno central detector,The European Physical Journal Plus139(2024) 1128

work page 2024

[2] [3]

L. Zhan, Y. Wang, J. Cao and L. Wen,Experimental Requirements to Determine the Neutrino Mass Hierarchy Using Reactor Neutrinos,Phys. Rev. D79(2009) 073007 [0901.2976]. [4]JUNOcollaboration,Calibration Strategy of the JUNO Experiment,JHEP2021(2021) 4 [2011.06405]

arXiv 2009

[3] [5]

Laser Calibration System in JUNO

Y. Zhang, J. Liu, M. Xiao, F. Zhang and T. Zhang,Laser Calibration System in JUNO,JINST14 (2019) P01009 [1811.00354]

work page Pith review arXiv 2019

[4] [6]

A.Takenakaetal.,Customizedcalibration sourcesinthe JUNOexperiment,JINST19(2024)P12019 [2410.01571]

work page Pith review arXiv 2024

[5] [7]

The Automatic Calibration Unit in JUNO

J. Hui et al.,The automatic calibration unit in JUNO,JINST16(2021) T08008 [2104.02579]

work page Pith review arXiv 2021

[6] [8]

Y. Guo, Q. Zhang, F. Zhang, M. Xiao, J. Liu and E. Qu,Design of the Guide Tube Calibration System for the JUNO experiment,JINST14(2019) T09005 [1905.02077]

work page Pith review arXiv 2019

[7] [9]

Y. Guo, K. Zhu, Q. Zhang, F. Zhang, Y. Meng, J. Liu et al.,Construction and Simulation Bias Study of The Guide Tube Calibration System for JUNO,JINST16(2021) T07005 [2103.04602]

work page Pith review arXiv 2021

[8] [10]

Cable Loop Calibration System for Jiangmen Underground Neutrino Observatory

Y. Zhang, J. Hui, J. Liu, M. Xiao, T. Zhang, F. Zhang et al.,Cable loop calibration system for Jiangmen Underground Neutrino Observatory,Nucl. Instrum. Meth. A988(2021) 164867 [2011.02183]

work page Pith review arXiv 2021