some aspects of the layout and optimization for the ... · some aspects of the layout and...
TRANSCRIPT
Some aspects of the layout and optimization for the cryogenic
supply of superconducting linacs
ERL2005 Workshop, Jefferson Lab,Va,USA, March 19-23,2005
Working Group 3
Bernd Petersen DESY, Hamburg, Germany
ERL2005 Bernd Petersen DESY
Overview• Introduction• High Q-Cavities Q = Q(T)• 2 K principle cooling scheme• Helium plant coefficient of performance COP = COP (T)• Cold Compressors• HeII Parameters = f(T)• Critical gas flow conditions = f (T)• Kapitza thermal resistance = f (T)• European XFEL-Linac as an example• Cryoplant parameters = f(T)• Conclusions
ERL2005 Bernd Petersen DESY
Introduction
ERL superconducting linacs will dissipate power
in the order of some kW at temperatures < 2.17 K
In view of conversion factors to primary power consumption
in the order of 1000 = primary power / load at low temperatures
it seems to be worth while to try some optimization
ERL2005 Bernd Petersen DESY
Introduction (cont.)
My talk is restricted to SL-linacs operated around 1.3 – 1.5 GHz
and to liquid helium II bath cooling below 2.17 K
The optimization is focused on the T< 2.17 K cooling circuit
( the 5-8 K / 40-80 K thermal shield niveaus are not discussed )The discussion is coupled to the believe that we can assume a
BCS Q = Q (t) dependence and the temperature dependent dynamicloads are dominating
DISCLAIMER:
Some XFEL-parameters, which are shown here, are in no way ‚official‘ project parameters. Often ‚hand-waving‘ argumentation isused to get qualitative results. Do not base the design of yourmachine on these data !
ERL2005 Bernd Petersen DESY
High Q Cavities
About factor 2 dynamic load reduction at 0.2 K temperature
reduction below 2.17 K !
1.00E+09
1.00E+10
1.00E+11
0 10 20 30 40
Eacc [MV/m]
Q0
2 K
1,8 K
1,6 K
1011
109
1010
AC70 1.3GHz 9-cell,TESLA-cavity
BCS T-dependence at medium fields
ERL2005 Bernd Petersen DESY
High Q requries proper magnetic field shielding !
AC70 result achieved by reduction of earth magnetic field to about 3 %
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Sensor position [m]
|B| [
G]
middlebottomtopLeftright
view from big s ide
natural field niveau: 0.25…0.3G
sm al s ide big s ide
Measurement of cryoperm magnetic shielding in horizontal cryostat
ERL2005 Bernd Petersen DESY
Q (T) is linked to the BCS surface resistance
Rs = A f2 exp –∆/kT (+ R0)
ERL2005 Bernd Petersen DESY
Diagramm taken from:
1.5 K is taken here as a reasonable lower limit of BCS Q=Q(T) dependence
Rs : BCS surface resistance ∆ : BCS energy gap A : material constant k : Boltzmann constantf : frequency T : temperature
R0: residual resistance
RS [nΏ]
TD-04-014, June 2004
2K Cooling Scheme ( simplified)
0.5 bar T cc out
Cold Compressors
P cc in T cc in
T bath = f ( P bath )
P bath
4.5 K load
40/80 K load
PrimaryPrimary PowerPower
HEX
HEX
HEX
HEX
JT
ERL2005 Bernd Petersen DESY
Coefficient of PerformanceCOP vs T
0200400600800
10001200
0 10 20 30 40Temperature [K]
CO
P
COP at Low T
0200400600800
10001200
1.5 1.7 1.9 2.1
Temperature [K]
CO
P
measured LHC value at 1.8 K
COP=1/( K * η CARNOT)
η CARNOT = T/(300 -T)
K= 0.176 ( from latest LHC measurements at 1.8 K)
COP= Primary Power/ Power at TT
TMR
JLab
ERL2005 Bernd Petersen DESY
Power Consumption
Refriger-ation
COP
W / W
Specific Load
% of Power
2 K 4.3 kW 588 2500 kW 49 %
5 – 8 K 7.5 kW 168 1254 kW 24 %
40 – 80 K 80.8 kW 17 1373 kW 27 %
Total 5147 kW 100 %
Power Consumption of the TESLA Model RefrigeratorExample for the required power at different T levels
ERL2005 Bernd Petersen DESY
Cold Compressor Cartridges of 2.4 kW @ 1.8 K Refrigeration Units
ERL2005 Bernd Petersen DESY
Cold compressor impellerIHI-Linde
The four-stage LHC cold compressors1st stage
Example of a cold compressorwith active magnetic bearingsused at Tore Supra, CEBAF and Oak Ridge
ERL2005 Bernd Petersen DESY
Development of the Efficiency of Cold
Compressors
Tem
pera
ture
(T)
Entropy (s)
P1
P2
1
2
2'
0
10
20
30
40
50
60
70
80
1985 1993 2000Year of construction
Isen
tropi
c ef
ficie
ncy
[%]
Tore SupraCEBAF
LHC
12
12'is HH
HHη−−
=
ERL2005 Bernd Petersen DESY
Cold Compressor Parameters
Insentropic efficiency of 0.7 assumed
Maximum compression ratio per single CC around 3 !
ERL2005 Bernd Petersen DESY
TEMP K
PRESS Pa
ENTROPY J/gK
ENTHALPY J/g
XFEL MASS FLOW g/s
KOMPRESSIONPOWER W
XFEL cw MASS FLOW g/s
KOMPRESSIONPOWER W
1.6 746 14.49 23.32
4 746 19.36 36.01
21.53 50000 19.36 126.8
28.99 50000 20.91 165.71 26.7 3462.99 135 17509.5
1.8 1638 13.43 24.2
4 1638 17.72 35.96
15.73 50000 17.72 96.54
20.7 50000 19.15 122.5 34.8 3011.592 264 22846.56
2 3129 12.58 25.04
4 3129 16.36 35.88
12.14 50000 16.36 77.7
15.56 50000 17.66 95.62 48 2867.52 466 27838.84
Helium II parameters = f(T)
Helium Parameters vs. Temperature
0
5
10
15
20
25
1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2
Temperature [K]
Hea
t of V
apor
isat
ion
[J/g
] / J
T Q
ual %
0.00
0.01
0.02
0.03
0.04
0.05
0.06
Vapo
r Pr
essu
re [b
ar]
Heat of vaporisation
JT-Quality %
Vapor pressure
-> more mass flow needed at LT
ERL2005 Bernd Petersen DESY
Helium II parameters = f(T) (cont.)-> Cooling instable below 1.8 K ? Test required !
Lambda, Cp,Diffusivity,Density of HEII
0
2000
4000
6000
8000
10000
12000
14000
16000
1 1.2 1.4 1.6 1.8 2 2.2
Temperature [K]
Lam
bda
[W/m
2], C
p [J
/kg]
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
Diff
usiv
ity [m
/s]
Lambda max in a saturated bath
Cp of liquid
'Diffusivity' = Lambda/(RHO*Cp) [m/s]
Liquid density=RHO [kg/m3]
ERL2005 Bernd Petersen DESY
Definition of Lambda max in a HEII bath
Heat conductivity in HeII q**m = f(T) * dT/dxT-Temperature, x-lengthm ≈ 3
f(T) Germany = f**(-1) (T) USAT2
q max * L** 1/3 = [ ∫ f(T) dT ]**1/3 T1
T1 = F ( P1 ) Temperature function of vapour pressureT2 = F ( P1 + ∆ P ) ∆ P = ρ*L*g ρ= density of liquid g=9.81 m/s2
L= depth in bath
T1, P1 at liquid surface
T2 must not beexceeded to avoidbubbles !T2, P2 at depth L in the bath
L
ERL2005 Bernd Petersen DESY
ERL2005 Bernd Petersen DESY
2.2 K forward
5 K forward
40 K forward
2 K 2-phase
cavity
2 K return
80 K return
8 K return
cool down/warmup
support
coupler
300 mm
XFEL-Cryomodule
2-phase flow
involved
2-phase flow characteristics for XFEL-Cryomodules
Flow pattern map for He II - vapor flow in the two-phase tube of 7 strings of TESLA XFEL (12,5 W / cryomodule)
1
10
100
1000
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
He II level / Tube diameter
K
Stratified Smooth - Wavy (K)
SSW_K (string 1)
SSW_K (string 4)
SSW_K (string 7)
( ) ( )βµρρρρ
cos
2
⋅⋅⋅−⋅⋅⋅
=LGL
LSGSGL
gUUK
Stratified Smooth
Stratified Wavy
TubeL
LLS A
mU⋅
=ρ
& TubeG
GGS A
mU⋅
=ρ
&
ERL2005 Bernd Petersen DESY
2-phase flow characteristics for XFEL-Cryomodules- > Lower densities balanced by smaller flows ?
K Values = f (T,m)
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
1 1.2 1.4 1.6 1.8 2 2.2
Temperature [K]
K-V
alue
0.00E+00
2.00E-01
4.00E-01
6.00E-01
8.00E-01
1.00E+00
1.20E+00
Gas
Den
sity
[kg/
m3]
Transition to wavy
4.8 g/s string mass flow for Q ( 2 K)
3.5 g/s string mass flow for Q ( 1.8 K)
2.7 g/s for Q(1.6 K)
Simplified dependencies : K ~ mG sqrt ( mL / ρL* ρG)
Density
ERL2005 Bernd Petersen DESY
Kapitza Heat Transfer-> Seems to be no limit under reasonable boundary conditions
0
50
100
150
200
250
300
350
400
1.5 1.7 1.9 2.1
Temperature [K]
Dis
sipa
ted
Hea
t [W
]
0
0.1
0.2
0.3
0.4
0.5
0.6
Kap
itza
Res
ista
nce
[W/K
*cm
2]
Hk=0.043*T**3.18 W/K*cm2
Heat transfer at the surface of oneTESLA cavity ∆ T = 100 mK
ERL2005 Bernd Petersen DESY
XFEL-Accelerator schematic layout
0.4
km1.
7 km
BC-I
BC-II
CollimationDiagnostics
Beam distr.
0.5GeV
10…20GeV (-->25GeV)
2 GeV
LINAC
Injector
beam lines
120 modules30 RF stations
Main linac Beam energy 20 GeV acc gradient 22.9 MV/m Bunch spacing 200 ns beam current 5 mA power beam p. klystron 3.8 MW incl. 10% + 15% overhead 4.8 MW matched Qext 4.6⋅106 RF pulse 1.37 ms Beam pulse 0.65 ms # bunches p. pulse 3250 Rep. rate 10 Hz Av. Beam power 650 kW
ERL2005 Bernd Petersen DESY
Site near DESY laboratory
ERL2005 Bernd Petersen DESY
3.2km
ERL2005 Bernd Petersen DESY
Wavelength vs. acc gradient
Photon wavelength vs E_acc, reference point 1 Angstroem at 17.5 GeV
00.20.4
0.60.8
11.2
1.41.6
15 17 19 21 23 25 27 29
E_acc [MV/m]
Lam
bda
[Ang
stro
em]
25 GeV17.5 GeV
Nominal linac energy 20 GeV, includes 57Fe line @ 0.8ÅExpected better cavity performance permits higher
energy/smaller wavelength
20 GeV XFEL parameters 10 Hz650 µs Pulse 23.5 MV/m
0
500
1000
1500
2000
2500
3000
3500
1 10 19 28 37 46 55 64 73 82 91 100 109 118Length [Cryomodules]
Vapo
r Pre
ssur
e [P
a]
0
0.0050.01
0.0150.02
0.025
0.030.035
0.040.045
0.05
Hel
ium
Mas
s Fl
ow ◘
kg/s○2K 1016 W
1.8 K 708 W
1.6 K 520 W
2K 48 g/s1.8 K 34.8 g/s
1.6 K 26.7 g/s
ERL2005 Bernd Petersen DESY
XFEL ERL-up-grade ?
0
500
1000
1500
2000
2500
3000
3500
1 9 17 25 33 41 49 57 65 73 81 89 97 105 113
Length of Linac [Cryomodules]
Vapo
r -Pr
essu
re [P
a]
0
0.1
0.2
0.3
0.4
0.5
0.6
Mas
s Fl
ow [k
g/s]
1.8K 5.373 KW
2K 10 KW
1.6K 2.618 KW
15 MV/m duty=0.5 Q=Q(T)
Q=Q(T)
mass flows
vapor pressure
vapor pressure
vapor pressure
ERL2005 Bernd Petersen DESY
All in one – the lower the better !?
ERL2005 Bernd Petersen DESY
0
200
400
600
800
1000
1200
1400
1600
1800
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 2.1 2.2
Temperature [K]
Prim
ary
Pow
er [W
]
0
0.5
1
1.5
2
2.5
Pow
er a
t Low
Tem
p [W
]
COP
CC
Work of compression * mass flow
COP + explicit CC
1 W dynamic
+1/3 W static (XFEL)
Conclusions
• Lowering the temperature seems to be effective as long as Q = Q(T) follows BCS and the temperature dependent dynamic loads dominate ( reasonable lower limit 1.5 K)
• Watch the temperature independent loads: HOMs, couplers…, static• The long term stability of the Q = Q(T) characteristic has to be proven• HeII cooling might become unstable below 1.8 K – tests required• Another cold compressor stage is required for each 0.2 K temperature step to lower
temperatures – investment costs and system complexity increase• In view of pressure drops, critical gas velocities, work of compression and general
sizing the lower gas densities at lower temperatures seem to be balanced by thelower cooling loads and the related lower mass flows
• If the cryogenic layout is suited for 2 K operation, it should also allow operation at lower temperatures as long as Q = (T) works ( beside the necessity of more CC stages) – for the time beeing this is our baseline for the XFEL
ERL2005 Bernd Petersen DESY