XFEL Cryo System Possible extension 1 st stage Project X Collaboration Meeting, FNAL September 8-9, 2010 (XFEL WP10 & WP13)
Outline 2 XFEL accelerator structure TESLA technology Basic cryogenic parameters Tasks & up-grade scenarios Heat load capacities of XFEL refrigerator Use of former HERA refrigerators Lay out of XFEL cryogenic system Cold turbo compressor issues Pressure variations/stability CW operation of TESLA cavities (reference to BESSY)
XFEL Photon Parameters 3
XFEL Accelerator requirements: 14 20 GeV electron beam energy About 642-962 superconducting RF cavities at 23 MV/m -> TESLA technology pulsed 10 Hz operation 4 Main linac: 1.65 km 80-120 cryomodules two independent injectors (1.3 GHz cm + 3.9 GHz cm each) Booster 12 cryomodules two bunch-compressor sections -> transferline bypasses needed 0.1 nm Photons X-ray FEL
XFEL-Accelerator Tunnel XTL 5
Main features of XFEL-cryomodule design based on the TESLA/TTF type III design -> use of TESLA -technology 6 10 Hz pulsed operation each cryomodule consist of: 8 1.3 GHz 9-cell Nb cavities (2K) 1 magnet package (2K) 2 thermal shields (5-8K;40-80K) 8 main RF couplers 8 cold tuners XFEL requirement: 23.6 MV/m Cryomodule cross-section
Accelerator Module (Cryomodule) 7 The XFEL accelerator module is based on the 3 rd cryomodule generation tested at the TESLA Test Facility and designed by INFN. Already 12 cryomodules have been built and commissioned for the TTF Linac. Length Total weight 12.2 m 7.8 t 38 carbon steel vessel 300 mm He gas return pipe acting as support structure 8 accelerating cavities cavity to cavity spacing exactly one RF wavelength inter-module cavity to cavity spacing a multiple of one RF wavelength one beam position monitor / magnet unit manually operated valves to terminate the beam tube at both ends longitudinal cavity position independent from the contraction / elongation of the HeGRP during cool-down / warm-up procedure 16
Accelerator Module (Cryomodule) 8 70 K shield 2 K return magnet current feedthrough 2.2 K forw ard 5 K forward 80 K return 8 K return RF main coupler 40 K forward 4 K shield 2 K 2-phase cav ity 17
Safety Measures 9 Safety flaps (up to DN300) installed on the vacuum vessel of each cryomodule LHC event : we reviewed helium release in XFEL tunnel
SC Magnet Package at 2K XFEL Magnet Package and Cold BPM 10 At the downstream end of the cavity string of each module a magnet package and an attached BPM is placed. a super-ferric quadrupole a vertical and a horizontal dipole BPM is either re-entrant (SACLAY design) or pick-up (DESY design) type. current leads Quadrupole to BPM alignment is 0.3 mm and 3 mrad. The magnet design is done in collaboration with CIEMAT. The current leads are based on the CERN design used at LHC. magnet package beam position monitor 2 K two-phase line 19
Choice of cryogenic parameters 11 2K Helium II bath: well below lambda transition close to heat conductivity maximum of saturated liquid Qo>10**10 for TESLA cavities (lower temperature has only limited advantage for pulsed RF operation) Low vapor pressure enables relative pressure stability in the order of less than +/- 1 % (problems with CCs!?) 5-8K circuit: 5K traditional from TESLA TDR (TESLA quad cooling,still main coupler thermal intercepts), fits to HERA cryo plant 40-80K circuit: Standard layout, fits to HERA plant
TESLA Cavities show BCS temperature dependence in Cryomodule 12 Module 6 CMTB Meas. Qo/Eacc average gradient 10Hz 500/800us Status:13-Mar-07 Esch/Kos/Lil/Lan MKS 1,00E+11 2K 19-Dec-06 2K 21-Feb-07 2K 28-Feb-07 with feedb/piezo 1.8K 1-Mar-07 1.55K 1-M ar-07 2K 9-M ar-07 1,8K 9-mar-07 1,59K 9-mar-07 Q o 1,00E+10 0 5 10 15 20 25 30 35 Eacc [MV/m]
XFEL-Linac-Cryogenic tasks (TDR) Very (NEW new start start ) version 13 642 816 944 sc Nb 1.3-GHz 9-cell cavities have to be cooled in a Helium II bath at 2K, Qo= 10**10, 23.6 MV/m, 10 Hz - - 40 3.9 GHz cavities (in 2 cryomodules) - 92 104 76 1,3 GHz cryomodules in RF operation - 812 4 cryomodules in cold stand-by - 2 Cryo-Bypass-Transfer-Lines (BCBTL) - at warm Bunch-Compressor sections - 2 independent injectors ( 1.3 GHz + 3.9 GHz cm each)
XFEL project realization in stages 14 Start version: 80 cryomodules 14 GeV Could be up-graded already during construction to 100 cryomodules 17.5 Gev Final up-grade (according to TDR): 120 cryomodules 20 Gev + cw injector
Heat loads & mass flows of final 20 GeV stage 15 13.4 KW 4.5K eq Subjects of process optimization
Heat loads & mass flows of 17.5 GeV stage 16 10.8 KW eq 4.5K
Operating conditions of XFEL Linac-Cryogenic 17 Continous operation of the refrigerator 24 h per day / 7 days per week Operation periods of 2 3 years without scheduled break of cold helium supply Avialability > 99% (without utilities and process control)
Use of HERA cryogenic plant 18 Preliminary studies from TU Dresden and a detailed industrial study by LINDE KRYOTECHNIK AG -> 2 HERA refrigerators (8 KW eq 4.5K each) could be modified to XFELrefrigerator ( + 2K system) Advantages: Only ½ of investment costs compared to new refrigerator No extra civil engineering effort Utilities exist Dis-advantages: Lower efficiency compared to new plant Re-use of worn equipment Parallel operation of 2 cold boxes challenging (symmetrical operation) There is still hope that only one CB will be sufficient to operate the 14 GeV linac ( or even the 17.5 GeV linac, if we ll not need all the capacity of the overhead)
Concept of XFEL-cryogenic 19 SC SC SC SC VC AMTF SC SC 2K BOX CC Valve BOX CB CB Distribution Box F CM VALVE BOX BCBTL BCBTL SCB 3.9 GHz CM F CMS CMS CMS CMS INJECTORS SCB BOOSTER MAIN LINAC
Main Transfer Line Piping Bridge to XSE 20
Cryogenic installations in XSE 21 Distribution Box
Concept of XFEL-Linac Cryogenic 22 Up-graded HERA plant Simplified flow scheme DB-box simplified 40K -> 80K shield circuit (in series) moved to XSE 5K -> 8K shield circuit (in series) 2K circuit (supply in parallel,return in series) warm gas collection pipe 40/80 K return, shield cooling 40/80 K forward 8 K return, shield cooling 5 K forward JTHEX 2.2 K forward 2 K return 40/80 K forward 5 K forward JTHEX 2 K return 40/80 K return, shield cooling lead cooling 8 K return, shield cooling 2.2 K forward Injector 2 (not shown) Injector 1 module string 9-10 module string 2 module string 1 cryogenic unit (9-10 strings), 1.7 km
Linac cryogenic string 23 Cryogenic units of 12 cryomodules = strings String Connection Boxes contain all cryogenic instrumentation Breaking news: All individual cavity helium vessels will get a thermometer!
Disadvantage: 2-phase flow affected by gravitational forces 24 Laser-straight XFEL-linac Liquid Helium II JT JT JT JT supply JT JT BC BC 8 Modules Length [m] Beam direction 66 141 208 354 442 String 11 String 10 String 9 One String = 12 cryomodules each of 12 m length Deviation from gravity equipotential [mm] 540 978 5 String 4 1561 1660 String 1 72,7 61,5 52,3 34,7 26,0 17,7 0,04 26,3 36,3 Gravity equipotential surface
Linac cryogenic components 25 regular string connection box End-BOX Bunch Compressor Bypass Transferline (only 1-phase helium) Cool-down/warm-up JT Feed-Box
Spec of Cold Turbo Compressors (CCs) 26 DESY MKS has NO experience with the operation of CCs We need advice from CERN,J-Lab, SNS Up to now, we learned from our CERN/LHC colleagues: - mixed-cycle approach mandatory - to increase CCs dynamic range - to restart from sub-atmospheric conditions of the linac - 5 stages approach too risky - make up of T,P, mass flow is a touchy operation
CCs mixed-cycle approach 27 DESY mixed-cycle approach: we could use the JT-screw (minimum 650 mbar) and/or make use of our helium pump units of the Accelerator Module Test Facility (AMTF) 2 X 20 g/s at 20 mbar (sufficient for static loads) 100 g/s at 200 mbar for mixed cycle + Extra heat exchanger in distribution box
2K reference specification ( simplified) 28 mixed cycle VC SC SC HEX HEX Existing Refrigerator Cold Compressors Details are subject of call for tender!? HEX JT JT subcooler HEX Critical for HERA CBs HEX JT T bath = f ( P bath ) 4.5 K load 40/80 K load
Sources of pressure variations: heavy single cavity quench in FLASH 29 Mass flow [g/s] P at helium pumps P at linac endcap Pressure [mbar] Temp [K] He mass flow T cavity He vessel 5 min!!!
XFEL pressure stability Spec 30 RF change of TESLA cavities vs pressure change : about 50 Hz / mbar Note from DESY RF experts: fluctuations not larger than +/- 35 Hz to avoid RF phase shifts -> pressure stability better than +/- 0.7 mbar required We specify 1% relative pressure stability -> 31 mbar +/- 0.3 mbar (LHC 5 % relative pressure changes caused by the accelerator components) Our question in call for tender: How does the CCs regulation react on a sudden mass flow change of 10%?
Pressure variations expected in XFEL linac 31 Arkadiy s questions: 1. Range of presures during normal operations? -> 31 mbar +/- 0.3 mbar ( 1%) 2. Timescale? -> Hours! From FLASH experience we expect very stable conditions 3. Occasional/rare pressure variations? -> Switching on/off RF, Quenches (see FLASH example) For the switching of RF we ll need some ramping by means of electrical heaters Trips of CCs?
CW-operation of TESLA cavities -> BESSY(1) W.Anders, J.Knobloch et al. at DESY: see SRF2009 paper D.Kostin, WD Möller et al. 32
CW-operation of TESLA cavities -> BESSY(2) 33
CW-operation of TESLA cavities -> BESSY(3) 34
CW-operation of TESLA cavities -> BESSY(4) 35
Summary 36 XFEL accelerator structure TESLA technology Basic cryogenic parameters Tasks & up-grade scenarios Heat load capacities of XFEL refrigerator Use of former HERA refrigerators Lay out of XFEL cryogenic system Cold turbo compressor issues Pressure variations/stability CW operation of TESLA cavities (reference to BESSY)