The Braidwood Reactor Neutrino Experiment Jonathan Link Virginia

The Braidwood Reactor Neutrino Experiment Jonathan Link Virginia Polytechnic Institute Workshop on Future PRC-U. S. Cooperation in High Energy Physics June 12, 2006 Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Reactor Experiment Basics νe νe νe Well understood, isotropic source of electron anti-neutrinos νe νe Oscillations observed as a deficit of νe Detectors are located underground to shield against cosmic rays. νe Probability νe 1. 0 sin 22θ 13 Unoscillated flux observed here πEν /2Δm 213 Distance (L) Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop 1500 meters

Braidwood Neutrino Collaboration 14 Institutions 70 Collaborators Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop 9/15/2020

Reactor Experiment Requirements The goal of next generation reactor experiments is to achieve sensitivity to sin 22θ 13 of better than 0. 01 (Current limit sin 22θ<0. 13 at 90% CL). To achieve this goal we need to… Reduce background rate to a few percent of neutrino signal in the far detector. Go deep underground & Reject backgrounds. Understand remaining background to less than a 0. 2% of signal. Tag background events for rejection and study. Know the relative normalization/efficiency of near and far detectors to better than 0. 3%. Movable detectors for cross calibration. Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Braidwood Design Principles Compare rate/shape in identical, large, spherical, on-axis detectors at two distances that have equal overburden shielding (Multiple detectors at each site: two near and two far) Systematic uncertainties cancel to first order and only have uncertainties for second order effects • Detectors filled simultaneously with common scintillator on surface • Large (65 ton target) detectors give large data samples • Spherical detectors to reduce any geometrical effects from neutrino direction and reconstruction • On-axis detectors eliminate any dependence on reactor power variations in a multi-reactor setup. • Equal overburden shielding gives equal spallation rates in near and far that can be exploited for detector and background checks Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop 9/15/2020

Braidwood Nuclear Station Located about 30 miles due south of the Fermilab. Owned and operated by Exelon Nuclear. • 2 Reactors with a total of 7. 17 GW thermal power • Well understood/favorable geology ─ Plant construction data ─ Site investigation (completed Jan. 05) • Strong support from owner • Ability to optimize layout Near Detector Jonathan Link, Virginia Tech Far Detector PRC/US Cooperation in HEP Workshop

Recent Geological Site Investigation Water Pressure Testing Audio Televiewer and Geophysics Suite Sand clay ………………. . 37' Carbondale Shale & Sandstone Maquoketa Shale ………………… 140' Maquoketa Limestone ………………… 207' Maquoketa Shale ………………… 271" Galena Limestone & Dolomitic Limestone NO TAKE Zones Depth 410 to 450 Depth 450 to 490 Depth 490 to 530 Depth 530 to 570 ………………… 467" Platteville Limestone & Dolomitic Limestone Depth 570 to 610 Depth 590 to 630 Jonathan Link, Virginia Tech GZA Geo. Environmental, Inc. PRC/US Cooperation in HEP Workshop

Background Radiation at Depth One of the standard geophysics probes measures the natural gamma radiation. It’s intended to measure rock porosity. Near Shaft Jonathan Link, Virginia Tech Far Shaft Confirms expectations that the background radiation in dolomitic limestone is relatively low. GZA Geo. Environmental, Inc. PRC/US Cooperation in HEP Workshop

Braidwood Baseline Design: Site Layout • Near detectors at a baseline of 285 meters (~4400 int/day) • Far detectors at a baseline of 1500 meters (~160 int/day) • All detector at a depth of 183 meters (600 ft) • Earth shielding is flat with an vertical cover of 464 mwe (as measured by our bore hole geological investigation) This should be compared to ~700 mwe cover under a mountain! Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Baseline Optimization The baseline optimization is sensitive to three effects: 1. Oscillation wavelength (πEν/Δm 2) 2. Event statistics which fall off like 1/L 2 3. The level of systematic error (S/N is a function of L which also varies as 1/L 2) Optimize in terms of the Δm 2 invariant, kinematic phase, and the percent of systematic error in terms of statistical error. Kinematic Phase = 1. 27Δm 2 L/E With Δm 2=2. 5× 10 -3 and systematic error at 150% of statistical error… L≈1500 meters Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Optimization for Shape vs. Rate Statistics Limited Systematic. Error= =200% 0% Systematic Error = 600% 1. The optimal Baseline for the systematics limited shape analysis is ~40º. 2. The optimal baseline for the systematics limited counting experiment is at the least optimal spot for a shape analysis! Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Braidwood Baseline Design: Detectors • Two zone design • 2. 6 meter target radius • 65 ton target • 3. 5 meter total radius 2. 6 m 3. 5 m Jonathan Link, Virginia Tech • 1000 PMT’s (coverage >25%) • 0. 2% Gd Loaded Scintillator Spherical design minimizes surface related BG and inefficiency, position dependence, buffer volume, and neutrino directional effects. PRC/US Cooperation in HEP Workshop

Two Zone Detector Design Two zone design offers simpler construction, optics, and source calibration, as well as larger fiducial mass for a given detector volume. Calibration from neutron capture peaks Large (r = 3. 5 m) detector reduces surface area to volume ratio, significantly reducing sensitivity to energy scale. 0. 1% uncertainty Use neutron capture peaks from inverse β -decay events to measure energy scale. • In each far detector, E scale can be measured to 0. 3% every 5 days. (This calibration averages over detector in exactly the same way as signal events. ) Acceptance uncertainty from energy scale in 2 zone design should be ~0. 1%. Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop 9/15/2020

Backgrounds The majority of backgrounds are directly related to cosmic rays (cosmogenic). There are three types of background 1. Random coincidence ─ where two unrelated events happen close together is space and time. 2. Fast neutron ─ where a fast neutron enters the detector, creates a prompt signal, thermalizes and is captured. 3. β+n decays of spallation isotopes ─ isotopes such as 9 Li and 8 He with β+n decay modes can be created in a spallation with μ on 12 C. Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Random Coincidence Background Assuming Kam. LAND concentrations of 40 K, 232 Th and 238 U and 450 mwe The rate of single events is shown on the left. Integrate plot in the positron and neutron signal regions to get the random rates. Positron-like: 142 k/det/day Neutron-like: 5070/det/day The ordered random coincidence of these events in 100 μs window is 0. 8/detector/day. Reverse order events give a handle on the background rate Increasing 238 U & 232 Th by orders of magnitude is a small effect. Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Fast Neutron Backgrounds There are three processes for the prompt “positron-like” events 1. Two neutron captures from the same cosmic ─ This should be tagged the vast majority of the time, but it sets the tag window for tagged muons at 100 μs. 2. Proton recoil off fast neutron ─ dominate effect. 3. Fast neutron excitation of 12 C ─ interesting, but not significantly different than 2. Energy spectrum peaks at particular values (like 4. 4 Me. V, first 12 C excited state) Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Tagging Muons at Braidwood The basic idea is to tag muons that pass near the detector so that we can reject the fast neutron background. Neutrons from farther away should mostly be ranged out. What eludes the tagging system? Shielding Veto Detectors 1. Veto inefficiency ─ 99% efficiency → 0. 25/detector/day p 2. Fast neutron created outside the shielding ─ 0. 5/detector/day n n 6 meters m Jonathan Link, Virginia Tech m With μ rate in the tag system of 21 Hz and the tag window of 100 μs → 0. 2% dead time PRC/US Cooperation in HEP Workshop

Comments on 9 Li Production Hagner et al. measured σ(μ+12 C->9 Li+X) assuming 9 Li was produced directly in μ-nucleon interactions: 9 Li However, it seems unlikely to produce 9 Li directly for Eμ>10 Ge. V. From our studies, it appears the production is a secondary of tertiary process: Deep inelastic scattering p n 9 Li 12 C(n, n 3 p)9 Li Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

What did Hagner Measure Anyway? Hagner measured 9 Li production in scintillator target cell from 190 Ge. V muons. Result: This is geometry dependent and hence not really a cross section! However, can use this to extract a cross section for 12 C(n, n 3 p)9 Li. Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Analog of Hagner Setup Using Braidwood G 4 MC 240 cm Rock Shield 20 million incident muons gives: 14 9 Li events in target cell 47 9 Li event in Z slice. 100 cm Z slice 88 cm Number of events in target Vessel cell corresponds to =2. 2 b consistent with Buffer Hagner Therefore using Hagner cross section underestimates the total number of 9 Li by a factor of three! Target Cell Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Fitting for 9 Li in Time Our simulations show that the vast majority of 9 Li production is associated with a muon that deposits >1 Ge. V in the detector. Time of Candidate Events since last μ > 1 Ge. V The 9 Li events have a half-life of 178 ms while the reactor event half life is determined from the rate of high deposited energy (showering) μ events. So the time distribution can be fit to two exponentials to determine the 9 Li contribution. Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Energy Spectrum Fit > 3 M events/near det Neutrino Energy ~3 M singles events/det Random Jonathan Link, Virginia Tech Fitting the energy spectrum takes advantage of the significant differences between the signal and backgrounds. Use tagged background events to measure the background energy spectra in the detector and then fit candidate spectrum for the untagged background contributions. ~12 K events/det Tagged Fast Neutron PRC/US Cooperation in HEP Workshop 9 Li 1800 events/det Tagged

Relative Normalization In the relative normalization many sources of uncertainty cancel. Remaining uncertainty must be less than 0. 3%. For a robust/believable result, need at least two independent ways to determine or constrain the remaining uncertainties Source of Relative Normalization Uncertainty Gd Fraction: Chemical Assay Observed Gd to H Capture Fraction (natural n only) Relative Detector Efficiency: Source Calibration Confirm Method with Cross Calibration Free Proton Count: Chemical Assay & Flow Measeurements Head-to-Head Cross Calibration Total Jonathan Link, Virginia Tech Error <<0. 1% 0. 3% <0. 2% 0. 3% 0. 3 -0. 5% PRC/US Cooperation in HEP Workshop 9/15/2020

Movable Detectors for Cross Calibration Spending just 8% of the run on head-to-head calibration (where two detectors sit next to each other in high flux of near site) results in a free proton calibration precision of better than 0. 3%. Detector calibration techniques can be verified against head-to-head neutrino relative calibration. Detectors are built and filled on the surface, facilitating a rapid deployment. νe νe νe Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Using Isotope Production As Addition Cross Check Unique feature of the Braidwood set up: Near and Far detectors at equal well-understood overburden Near and Far detectors have substantial shielding Can use produced 12 B events to check: Near/far relative target mass from the total rate Near/far energy calibrations from the relative energy distribution 12 B Beta decays t 1/2 = 20 ms (can tag to muon) 13. 4 Me. V endpoint ~50, 000 12 B beta-decay events per year per detector can be tagged and isolated for a statistical uncertainty of 0. 45% Systematic uncertainties related to the relative near/far overburden needs to be known to few percent from: Geological survey information (Bore hole data: near/far agreement at <1%) Cosmic muon rates in the near and far locations Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop 9/15/2020

Braidwood Projected Sensitivity For three years of Braidwood data and Δm 2 ≈ 2. 5× 10 -3 e. V 2 90% CL limit at sin 22θ 13 < 0. 005 3 σ discovery for sin 22θ 13 > 0. 013 Summary of Uncertainties for a 3 year Data Run Sensitivity to sin 22θ 13 Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Conclusions The measurement rests on the ability to understand the near/far relative normalization and to mitigate and measure backgrounds. Each contribution to the relative normalization should be measured and verified by at least two independent methods. This redundancy is critical to ensure that the high precision goals of the experiment can be meet. Background reduction is achieved by tagging muons and shielding the detector from spallation products of untagged muons. 9 Li is tagged by muon energy deposition in the detector. Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Conclusions All background sources can be fit for using energy spectra derived from the tagged samples. The overall three year sensitivity goal was sin 22θ 13< 0. 005 at Δm 2 = 2. 5× 10 -3. Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Question Slides Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop

Detectors Designed to Move 9 meters Detector The detectors need to carry electronics and front-end DAQ as these components also play a role in detector efficiency. Multiple far detectors may be used to maintain portability while increasing total volume. Detectors need to fit in the shaft (Diameter ≤ 7 meters) 8 meters Jonathan Link, Virginia Tech PRC/US Cooperation in HEP Workshop ►