The Spallation Neutron Source Project Geneva December 8, 2006

Transcription

1 The Spallation Neutron Source Project Geneva December 8, 2006 Norbert Holtkamp ITER International Organization St Paul lez Durance

2 The Spallation Neutron Source SNS is funded through DOE-BES and has a Baseline Cost of 1.4 B$ 1.3 GeV facility designed & build to operate at 1 GeV to begin with SC linac Single ring capable to operate at 1.3 GeV One target station The peak neutron flux will be ~20 100x ILL SNS has begun operation with 3 instruments installed. 2

3 The SNS Mega Terms What it is and what it is supposed to be in a few years It is: The first high energy proton linac largely built with superconducting RF structures (0.812 GeV out of 1.0 GeV). The worlds highest energy proton linac (operated with H-) The second largest accelerator RF installation in the US. The first Multilab collaboration with fully distributed responsibility for accelerator construction. A project that finished On Time and within budget according to a schedule/budget set in It will have: The highest intensity proton storage ring of its kind. The highest average beam power available in the world. The most advanced Neutron scattering facility with the best in class instruments 3

4 Spallation-Evaporation Production of Neutrons and Why to use heavy metal target! Fission chain reaction continuous flow 1 neutron/fission Pbeam E ev I A Tpulse sec f rep Hz SNS : Goal is to achieve MW average power Spallation no chain reaction pulsed operation 30 neutrons/proton Time resolved exp. 4

5 18 Instruments Now Formally Approved (20+ funded) Fundamental Physics to Chemistry to Genomes to Life Engineering 5

6 Backscattering Spectrometer 84 m incident flight path designed to provide high energy resolution 2.5 mev (fwhm) at the elastic line slow dynamics (100 s psec, 3 35 Å) Approximately 50 x faster then current world s best comparable instruments better Q-resolution simplifies studies involving crystalline materials 6

7 Reactors vs. Accelerator-Driven Sources Reactor-based source: neutrons produced by fission reactions Continuous neutron beam 1 neutron/fission Accelerator based source: Neutrons produced by spallation reaction 25 neutrons/proton for Hg Neutrons are pulsed and follow proton beam time structure A pulsed beam with precise t0 allows neutron energy measurement via TOF 7

8 Spring

9 Spring

10 SNS Accelerator Complex Front-End: LINAC: Accumulator Ring: Produce a 1msec long, chopped, Hbeam Accelerates the beam to 1 GeV Compress 1 msec long pulse to 700 nsec H- stripped to protons RFQ DTL Current 945 ns 186 MeV CCL 1000 MeV 387 MeV SRF, b=0.61 SRF, b=0.81 Chopper system makes gaps mini-pulse 1 ms macropulse Current 2.5 MeV 87 MeV Ion Source Deliver beam to Target 1ms 10

11 Target Region Within Core Vessel Target Module with jumpers Outer Reflector Plug Target Moderators Core Vessel water cooled shielding Core Vessel Multi-channel flange 11

12 Original SNS CDR for CD-1 May MW 1.0 GeV copper CCL linac with no room for increased energy, only current HEBT and Ring magnets sized for 1.0 GeV H- Orginal Design has very little to do with what was built. 12

13 SNS High Level Baseline Parameters Beam Energy Average Beam Current Beam Power on Target Pulse Repetition Rate Beam Macropulse Duty Factor H- Peak Linac Current Linac Beam Pulse Length Ring Beam Extraction Gap Protons Per Pulse on Target Proton Pulse Width on Target Uncontrolled Beamloss Criteria Linac length Total Beamline Length Target Material Energy Per Beam Pulse Maximum Number Instruments GeV 1.4 ma 1.4 MW 60 Hz 6.0 % 38 ma 1.0 ms 250 ns 1.5x ns 1 Watt/m 335 m 903 m Liquid Hg (1 m3; 18 tons) 24 kj 24

14 The Spallation Neutron Source Partnership Description Project Support Front End Systems Linac Systems Ring & Transfer Systems Target Systems Instrument Systems Conventional Facilities Integrated Control System BAC Contingency TEC R&D Pre-Operations TPC Accelerator , , ,411.7 SNS-ORNL Accelerator systems: ~167 M$ ~20 M$ ~113 M$ ~106 M$ ~177 M$ ~60 M$ At peak : ~500 People worked on the construction of the SNS accelerator ~63 M$ The partners developed and built SNS/ORNL integrated, installed + operated 15

15 SNS Multilab Organizational 17

16 SNS CD-4 Scope Criteria Level 0 DOE Deputy Secretary Accelerator-based neutron scattering facility capable of at least 1 megawatt proton beam power on target Level 1A Director, Office of Science Five specific research instruments Level 1B Associate Director BES Performance test demonstrating 1.0E+13 protons per pulse (one pulse is enough) 5.0E-3 neutrons/steradian/incident proton, viewing moderator face (~x5-10 under normal moderator performance) Level 2 DOE Project Director At least 3 of the 5 instruments installed and tested; the other 2 procured and on site Trained staff, operating permits, and systems documentation in place 18

17 Budget Driving the Schedule DOE supported SNS immensely by making sure that we got the budget to execute the plan that was laid out. The schedule had a very aggressive procurement plan, that was based on the idea of an accelerator in the box 19

18 Cost Development between 2001 and 2006 Spend $1.41 Billion dollars in 7 years with a peak of ~ 1 M$/day during peak construction. ~ $6.5 M contingency left at the end for scope additions Operations budget 160M$/Year 450 staff positions ~2000 users per year at full capacity. Nov 2001 [$M] May 2006 [$M] Contingency 1.01 Research & Development % 1.10 Operations % Total OPC (Burdened, Escalated Dollars) Project Support % 1.03 Front End Systems % 1.04 Linac Systems % 1.05 Ring & Transfer System % 1.06 Target Systems % 1.07 Instrument Systems % % % % 1.08 Conventional Facilities 1.09 Integrated Control Systems Total Line Item (Burdened, Escalated Dollars) %

19 LBNL: SNS Front-End Systems Front-End H- Injector was designed and built by LBNL Front-end delivers 38 ma peak current, chopped 1 msec beam pulse H- Ion Source has operated at baseline SNS parameters in several endurance runs since Ion Source Radio-Frequency Quadrupole >40 ma, 1.2 msec, 60 Hz 21

20 LANL: Normal Conducting Linac & RF Systems MHz DTL was designed and built by Los Alamos Six tanks accelerate beam to 87 MeV System includes 210 drift tubes, transverse focusing via PM quads, 24 dipole correctors, and associated beam diagnostics CCL Systems designed and built by Los Alamos 805 MHz CCL accelerates beam to 186 MeV System consists of 48 accelerating segments, 48 quadrupoles, 32 steering magnets and diagnostics 22

21 DTL/CCL Commissioning Results Full transmission of accelerated beam to the beamstop (with few % measurement uncertainty) Typical beam pulse: 20 ma, 40ms, 1 Hz (limited by intercepting diagnostics and beamstop) Module Design [MeV] Measured [MeV] Deviation [%] DTL ± CCL ± CCL ± CCL ± July 2004

22 JLAB: The Superconducting Linac Designed and built by Jefferson Laboratory. SCL accelerates beam from 186 to 1000 MeV. SCL consists of 81 cavities in 23 cryomodules. Two types of cavity geometries are used to cover broad range in particle velocities (β=.65,.85). Cavities are operated at 2.1 K 1. with He supplied by Cryogenic Plant. 2. Operation so far mostly at 4.2 K Superconducting RF Advantages: Flexibility gradient and energy are not fixed More power efficient lower operational cost High cavity fields less real estate Better vacuum less gas stripping Large aperture less aperture restrictions reduced beam loss reduced activation 24

23 Representative Linac Beam Pulse 860 MeV 18 ma peak current 200 msec 70% Chopping 12 ma average pulse current 1.5x1013 H/pulse Overlay of 12 linac beam current monitors August

24 Cavity Gradient Performance 16 Maximum fields achieved in the installed cavities. 2.1 K Open loop 4.2 K Open loop Number of Cavities Operational fields are kept in general at 75-80% of the maximum fields 2.1 K Closed loop Ea [MV/m]

25 Energy Stability Pulse to Pulse 865 MeV beam ~ 1000 pulses 20 msec pulse 12 ma beam RMS energy difference jitter is 0.35 MeV, extreme = MeV (without any energy feedback) Parameter list requirement is max jitter < +1.5 MeV 27

26 Low-level RF Feedforward within the Beam Pulse Beam turn-on transient gives RF phase and amplitude variation during the pulse, beyond bandwidth of feedback. LLRF Feedforward algorithms have been commissioned. Without Feed-forward With Feed-forward 28

27 Linac RMS Transverse Emittance Measured e (H, V) norm. p-mm-mrad Parameter List e (H, V) norm. p-mm-mrad Notes MEBT Entrance 0.22, RFQ Exit Twiss study CCL Entrance 0.22, Matching 7 CCL profile sets SCL Entrance 0.27, Matching 3 SCL profile sets Linac Dump 0.26, wire, vary quads Measured RMS emittance is within specification but beam parameters are different for various runs. 29

28 Linac RF Systems Designed and procured by LANL All systems 8% duty factor: 1.3 ms, 60 Hz 7 DTL Klystrons: 2.5 MW MHz 4 CCL Klystrons: 5 MW 805 MHz 81 SCL Klystrons: 550 kw, 805 MHz 14 IGBT-based modulators 81 SCL Klystrons High Voltage Converter Modulators 2nd largest klystron and modulator installation in the world! DTL Klystrons 30 CCL Klystrons

29 JLAB: SNS CHL Facility 31

30 Warm Compressors System Status 3 warm compressor streets 32

31 4.5K Cold Box has been Commissioned 8 kw at 4.5 K 33

32 JLAB + ORNL: The SNS Cryogenic Support System ~2 2.0 K ( 40mBar) Operated since Oct 2004 uninterrupted. Mainly at 4.5K Very reliable and build to accommodate an additional 9 (23->31) cryomodules. 34

33 BNL: Accumulator Ring and Transport Lines Collimation Accumulator Ring Circum Energy frev Qx, Qy Qx,y x, y Accum turns Final Intensity Peak Current RF Volts (h=1) (h=2) 248 m 1 GeV 1 MHz 6.23, 6.20 RF , x A 40 kv 20 kv Extraction Injection RTBT HEBT Target 35

34 Ring and Transport Lines HEBT Arc Ring Arc Injection RTBT/Target 36

35 Instrument floor layout can be seen and target installation began 37

36 The SNS Target: 2-MW Design 25 kj/pulse at 7x15cm beam size sets of transverse and longitudinal shock wave. Needs to be exchanged remotely every 3 month at full power operation. 1 mm Back of Target Service Bay 38

37 Recent Achievements Ring commissioning started Jan 12th, Status on January 14th in the picture. Quick start up of charge exchange injection, accumulation, and extraction since all diagnostics online at commissioning begin. 7 turn injection Extraction beam dump 39

38 Push For High Intensity Achieved CD4 1.3 x 1013 ppp intensity on Jan 26. multiturn injection Extraction beam dump 40

39 High Intensity Results: Beam Loading in the Ring RF System 3x1013 protons per pulse Extraction 5x1013 protons per pulse 41 Shows distortion of longitudinal profile and beam leaking into gap due to untuned compensation of beam loading in RF

40 Instability Studies at Up to 1014 Coasting Beam During high-intensity studies we searched for instabilities by delaying extraction operating with zero chromaticity storing a coasting beam No instabilities seen thus far in normal conditions First instability observed with central frequency about 6 MHz, growth rate 860 us for 1014 protons in the ring, Scaling these observations to nominal operating conditions predicts threshold > 2 MW Fast: electron-proton Slow: Extraction Kicker Evolution of 75th Harmonic 6 Log(maginitude(75th Harmonic)) Turns

41 1.6x1013 Protons Delivered to the Target for CD4 Beam Demonstration (April 28, 2006) Ring Beam Current Monitor Final RTBT Beam Current Monitor 43

42 Beam on Target: Injection Painting Beam on Target View Screen Beam profiles in RTBT 65 mm 80 mm 44

43 Ring/RTBT/Target Commissioning Timeline January-May 2006 Jan. 12: Jan. 13: Jan. 14: Jan. 15: Jan. 16: Jan. 26: Feb. 11: Feb. 12: Feb. 13: Received approval for beam to Extraction Dump. First beam to Injection Dump. First beam around ring. >1000 turns circulating in ring First beam to Extraction Dump. Reached 1.26E13 ppp to Extraction Dump. ~8 uc bunched beam (5x1013 ppp) ~16 uc coasting beam (1x1014 ppp) End of Ring commissioning run April 3-7: Readiness Review for RTBT/Target April 27: Received approval for Beam on Target April 28: First beam on target and 2 hours later CD4 (>1013) beam demonstration 45

44 Primary Concern for SNS on its Way to Full Power Operation: Uncontrolled Beam Loss Hands-on maintenance: no more than 100 mrem/hour residual activation (4 h cool down, 30 cm from surface) 1 Watt/m uncontrolled beam loss for linac & ring Less than 10-6 fractional beam loss per tunnel meter at 1 GeV; 10-4 loss for ring High rad areas Uncontrolled loss during normal operation Beam loss [W/m] FE 1.5 DTL 1 CCL 0.5 RING HEBT SCL RTBT Length [m]

45 SNS Diagnostics Deployment Operational Not Operating MEBT 6 Position 2 Current 5 Wires 2 Thermal Neutron 3 PMT Neutron 1 fast faraday cup 1 faraday/beam stop D-box video D-box emittance D-box beam stop D-box aperture Differential BCM IDump 1 Position 1 Wire 1 Current 6 BLM RING 44 Position 2 Ionization Profile 70 Loss 1 Current 5 Electron Det. 12 FBLM 2 Wire 1 Beam in Gap 2 Video 1 Tune EDump 1 Current 4 Loss 1 Wire CCL 10 Position 9 Wire 8 Neutron, 3BSM, 2 Thermal 28 Loss 3 Bunch 1 Faraday Cup 1 Current DTL 10 Position 5 Wire 12 Loss 5 Faraday Cup 6 Current 6 Thermal and 12 PMT Neutron SCL 32 Position 86 Loss 9 Laser Wire 24 PMT Neutron CCL/SCL Transition 2 Position 1 Wire 1 Loss 1 Current 47 RTBT 17 Position 36 Loss 4 Current 5 Wire 1 Harp 3 FBLM HEBT 29 Position 1 Prototype Wire-S 46 BLM, 3 FBLM 4 Current LDump 6 Loss 6 Position 1 Wire,1 BCM

46 Losses in HEBT and Ring HEBT Inj Dump Collimators 480 turn accumulation Relative Idump/Ring losses reduced with smaller spot on foil 48 Ring

47 SNS Early Operations: Ramping up Scientific Productivity Shared the plan with the community to get them involved early on: Manage Expectations Ac c e le ra tor Av a ila bility a nd Ope ra tion % Ops Hours This slide was made on plane ride in 2 h and most of the numbers were invented U s er Ops hours A c c el. P hy ic s B eam P ow er/k W R eliability (% ) Ye a rs U s er Ops hours A c c el. P hy ic s 49 B eam P ow er/k W R eliability (% )

48 Schedule Changes. And How Did We Make It? Its always the first schedule that counts to measure how well a project is doing, not the last one.. FY actual 240 days Front-End DTL/CCL SCL DTL Tank 1 The End Ring DTL Tanks 1-3 Target 2001 plan 460 days 50

49 How Much R&D Can One Do/Effort in a Construction Project? Quite a bit and a construction schedule is driving the R&D to be very fast and efficient. LASER profile monitor to replace standard carbon wires in the SC part of the linac which can be used while operating a full intensity beam Nano-crystaline foil development for high intensity beams, tested at the PSR. A fast feedback system to reduce/eliminate the PSR instability A H- stripping experiment based on Laser/Magnetic stripping. 51

50 Where Does SNS Stand today?: Performance Goals 95 Beam Power Goal Neutron Production Hours Reliability FY07 12 FY08 Months FY Reliability Beam Power (kw), Production Hrs 2500

51 Operations Planning: Ramp-up in Beam Parameters Rep. Rate (Hz) Peak Current (ma) Pulse Length (u-sec) Power (kw) Oct06 Apr- Oct07 07 Apr08 Oct08 Pulse Length ( m sec) / Power (kw) RR (Hz) / Current (ma) Power Ramp Up Apr- Oct09 09 Beam repetition rate is consistent with the operations envelope limit 53

52 Beam Power [0-60 kw] Beam-Power-on-Target History Beam power administratively limited to 10 kw until November 8 May 1, 2006 Nov 30,

53 Run : Integrated Beam Power by Day and Cumulative 6.3 MW-hrs delivered in Run

54 Run Statistics Run Parameters: Beam Energy: 890 MeV 20 kw: 5 Hz rep-rate, 300 msec linac beam pulse, 20 ma peak current kw Hrs per week Neutron Production Hours per Week /1 9/ /1 2/ /5 / /2 9/ /2 2/ /1 5/ /8 /2 0 Hours/Week 70 kw Hrs/Week 90

55 Ramp-Up Progress Peak Beam Power kw-hrs/day Weeks Since First Beam on Target Peak Beam Power (kw) Average kw-hrs/day

56 Run Breakdown Statistics Pulsed power: modulators and beam choppers Accelerator de-ionized water systems Mercury pump 58

57 FY07 Accelerator Improvement Projects Project Description FY07 High-Voltage Converter Modulator Upgrade Phase 1 Upgrade HVCM systems to improve reliability and rep rate; install modulator test stand 1,450 Accelerator Cooling Water System Reliability Upgrade Upgrade water systems to improve reliability 1,150 Improve diagnostics instrumentation to reduce beam loss and component activation 1,300 Beam Diagnostic Upgrade Phase 1 LEBT Chopper Upgrade MEBT Chopper Upgrade RF System Upgrade Phase 1 Accelerator Power Supply Lock-out Tagout System CHL Tie-in to Test Cave Injection Region Upgrade Phase 1 Redesign LEBT chopper system with greater engineering margin to improve reliability 400 Redesign MEBT chopper system with greater engineering margin to improve reliability 230 Upgrade MEBT RF systems and RFQ couplers for higher reliability; other reliability improvements 1,400 Install trap-key system and finger-safe terminals for HEBT/Ring/RTBT power supply systems to enable rapid and safer lock-out/tag-out 400 Connect RF test cave to CHL; add permanent, movable shielding door to cave 770 Incorporate additional corrector magnet, vacuum chamber and diagnostics in injection dump line; install optimized primary foil mechanism to reduce beamloss 900 Total AIP $8,000k 59

58 Run : Goals vs. Achieved 90 Integrated Beam Power [MW-hrs] Quarterly Integrated Beam Power Goal Actual Integrated Beam Power Three goals for this13run: 26 Weeks Deliver > 3.9 MW-hrs in neutron production Demonstrate sustained 30 kw operation Demonstrate 60 kw capability for > ½ shift

59 Summary SNS had a very rough start in the 90 s when it was converted from a fission reactor project to an accelerator based neutron source to a sc linac driven source. (each change was accompanied by a new management team) The SNS project has officially finished at the end of May, 2006 (end of June 2006) with the signature of the Critical Decision 4 documents at a total project cost of $1,405,2M ($1,411.7M). SNS construction was accompanied by several severe technical setbacks and subsequent MIRACLES on recovery. This is probably my last official talk about SNS and it s a pleasure to give it CERN here in the lecture in front of such a distinguished group. Thanks 61

60 Technical Challenges: Beam Loss The SNS is a loss-limited accelerator; losses must be kept < 1 W/m to limit residual activation We measure higher than desired losses in a few locations: Ring Injection region We are unable to simultaneously transport waste beams (from stripping process) to the injection dump and properly accumulate in the ring Short-term fixes allow >100 kw operation; midterm fixes (April 2007) are in preparation; longterm fix requires redesign of injection dump beamline and 2 new magnets Internal Review of Injection Dump performance and recovery options was held Nov. 21st with R. Macek from LANL (D. Raparia from BNL will review results in December) Coupled cavity linac due to orbit control and sparseness of loss monitors Near RTBT Dipole for large painted beams Active and aggressive accelerator62 physics studies have reduced

61 Technical Challenges: Superconducting Linac Performance Superconducting Linac is operating at 892 MeV All but 3 (of 81) cavities are operational at low rep-rate (< 5 Hz) As a conservative measure, we have turned off 6 additional cavities in order to confidently operate at higher rep-rates (15 Hz) Driven by concern over potential Higher-Order-Mode feedthrough failures; certain cavities show HOM waveforms that indicate pathological behaviour A plan which follows the 3-year beam power ramp-up curve is being formulated: We are working with Jefferson Laboratory to develop plans for Reworking one high-beta cryomodule (remove in December) Reworking the medium-beta prototype cryomodule to turn it into an operational spare We are preparing a bid package for procurement of spare high-beta cryomodules We are establishing cryomodule repair, maintenance and testing capabilities on-site (AIP Project) Cleanrooms are on order 63

62 Technical Challenges: Reliability Limitations Beam Chopper Systems Repeated failures in LEBT and MEBT chopper systems New, more robust, designs will be manufactured this year (FY07 AIP) High-Voltage Converter Modulators A number of weak components limit MTBF to 2700 Hrs Several prototype improvements are in test in single operational units Begin upgrade program for better fault detection, replacement of components with higher engineering margin, (FY07 AIP) Water Systems Flow restrictors continue to clog Responded to floods from failed gaskets Have been replacing all gaskets and retorquing flanges in klystron gallery, service buildings and tunnel Will remove flow restrictors during December down period, replace with valves where necessary Water systems upgrade program included in FY07 AIP plan. Cryogenic Moderator System Thermal capacity degrades-- requires cycling every 2 weeks; Manufacturer will attempt repair in December Mercury Pump Seal failed Nov. 26; repair strategy being formulated 64

63 DTL and CCL RF Setpoints by Phase Scan Signature Matching BPM Phase Diff (deg) J. Galambos, A. Shishlo CCL Module 2 RF Phase 65

64 SCL Phase Scan using BPMs BPM phase diff SCL phase scan for first cavity Solid = measured BPM phase diff Dot = simulated BPM phase diff Red = cosine fit Cavity phase Matching involves varying input energy, cavity voltage and phase offset in the simulation to match measured BPM phase differences Relies on absolute BPM calibration With a short, low intensity beam, results are insensitive to detuning cavities intermediate to measurement BPMs 66

65 Ring Closed Orbit: H,V Bumps are Due to Injection Kickers Horizontal Orbit Vertical Orbit BPM Amplitude 67

66 Turn by Turn Data for Zero Chromaticity Horizontal Turn-by-turn Vertical Turn-by-turn Amplitude Turn-by-turn 68

67 Ring Optics Measurements: Betatron Phase Advance and Chromaticity Plots show measured betatron phase error vs. model-based fit Horizontal Fit Betatron Phase Error (degrees) Horizontal Data Horizontal Fit -20 Data indicates that the linear lattice is already very close to design -30 Beam Position Monitor Vert Horiz Natural Chromaticity (Design) Natural Chromaticity (Measured) Corrected Chromaticity (Meas) Vertical Fit Betatron Phase Error (degrees) Vertical Data Vertical Fit Beam Position Monitor 69

68 SCL Laser Profile Measurements Measured horizontal profile after SCL cryomodule 4 SCL laser profiles (H + V) were available at 7 locations 3 at medium beta entrance, 3 at high beta entrance and 1 at the high beta end Expect reliable data beyond 3 sigma during operation. 70

69 E-P Feedback Experiment at the PSR We formed a collaboration to carry out an experimental test of active damping of the e-p instability at the LANL PSR: We deployed a transverse feedback system designed and built by ORNL/SNS and in two shifts demonstrated for the first time damping of an e-p instability in a long-bunch machine In subsequent studies we observed a 15-30% increase in e-p instability threshold with feedback on. Continued investigation of e-p feedback will be pursued, as well as simulations to benchmark experimental results. 71

70 Example Measurements Position, phase Current (toroids) Loss (ion chambers) emittance Current (MEBT beam stop) Current (D-plate beam stop) Current (DTL Faraday cup) Loss (neutron) Profile (wires) Halo 72 Bunch shape

71 Laser-Stripping Injection Proof-of-Principle Experiment We are receiving funds from the Lab Directors R&D (LDRD) program to perform a proof-ofprinciple experiment to test a scheme for laser-stripping injection First experimental tests were carried out in early December We observed >50% doublestripping efficiency in first attempt in a one-hour run Further R&D will continue, with the goal of developing a realistic laser-based scheme Flipped-sign notch on BCM indicates protons Laser Beam High-field Dipole Magnet H- H0 Step 1: Lorentz Stripping H- H0 + e- 73 H0* Step 2: Laser Excitation H0 (n=1) + H0* (n=3) High-field Dipole Magnet proton Step 3: Lorentz Stripping H0* p + e-

72 Now he should leave Thanks for all the congratulations!!!! 74