1. LLNL A Sandia and Lawrence Livermore National Laboratories Joint Project Nathaniel Bowden Detection Systems and Analysis Sandia National Laboratories, CA Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy under contract DE-AC04-94AL85000. The Safeguards Detector at SONGS

2. LLNL Design Principles Simple, inexpensive, robust –Rapid deployment –Use well known detection concepts/technology Antineutrino detection via inverse beta decay Gd loaded scintillator central target surrounded by various shielding layers –Physically robust for reactor environment (e.g. steel scintillator vessels) –Modular for manhole access Do a relative measurement –Use automatic calibration based on background lines to account for all time dependent variations

3. LLNL Sandia/LLNL Antineutrino Detector Detector system is… –0.64 ton Gd doped liquid scintillator readout by 8x 8” PMT –6-sided water shield –5-sided active muon veto

4. LLNL Cell Design Stainless tanks – no scintillator attack –Tank size determined by manhole size PMTs coupled to scintillator by acrylic plugs and mineral oil Light reflectors are argon filled PTFE bags (Bugey)

5. LLNL Prototype deployment – San Onofre Nuclear Generating Station

6. LLNL Tendon gallery is ideal location –Rarely accessed for plant operation –As close to reactor as you can get while being outside containment –Provides ~20 mwe overburden 3.4 GWt => 10 20 / s In tendon gallery with ~10 17 / s per m 2 Around 4000 interactions expected per day San Onofre Nuclear Generating Station Unit 2 Tendon Gallery

7. LLNL Installation at SONGS

8. LLNL Installation at SONGS

9. LLNL Some results Detector is ~ 10% efficient Stability is difficult to maintain with only background lines for calibration Even so, reactor power excursions are clear; probably burnup too

10. LLNL Background singles rate is high With hardware threshold at ~ 1.5 MeV, singles rate is ~500 Hz Analysis threshold is 3 MeV

11. LLNL Our detector is “all edge” A large fraction of the  -rays from the Gd shower escape our detector, resulting in a broad delayed energy distribution Data Monte Carlo Events/MeV Delayed Energy (MeV)

12. LLNL Lessons Learnt We need: –Better gamma shielding/cleaner material –More, and more uniform, light collection –Better calibration (background lines won’t be enough, no sources possible?) We would like –Smaller footprint –Less flammable/aggressive scintillator –Smaller surface/volume ratio Leading to higher efficiency in a smaller volume, with excellent stability

13. LLNL  13 vs. nonproliferation? “State of the Art” vs. a detector that is “good enough”

14. LLNL Conclusions Our very simple device has made interesting measurements and has been invaluable as a demonstration, but we can and must do better We are likely to begin a new detector development program this year, beginning by studying the use of steel shielding with shallow overburden It is important in our discussions to identify the necessary features to make nonproliferation detectors successful, but not too complex or expensive