1. Status of studies on FCC magnet circuit architecture and protection
Marco Prioli
CERN, TE-MPE
FOR

2. Magnets parameters
All numbers in this presentations are obtained for the following magnet versions
Load line margin
Bnom [kA]
Differential inductance (2 apertures) [mH]
Stored Inom (2 apertures) [MJ]
Cos-theta v22b
14%
11.23
566 38
Block coil v20ar
10.99
571 36
Common coil v1h2_1ac1
16.80
287 43
Canted cos-theta 18
254 44
Cos-theta v22b
Block coil v20ar
Common coil v1h2_1ac1
Canted cos-theta

3. Timeline 2 presentations about circuit protection t
April 2016
FCC week
November 2016
ECC WP5 review
May 2017 t
2 presentations about circuit protection
M. Prioli, “Concepts for magnet circuit powering and protection” (Link)
A. Verweij, “Comparison of magnet designs from a circuit protection point of view” (Link)

4. 1 - FCC vs LHC powering layout
In order to reduce the circuit inductance, the half arc of the FCC is subdivided in two powering sectors (PS)
The total number of power converters (PC) is doubled
LHC
FCC
PC 1
PC 2
PS 1
PS 2
HalfArc 1
8km
PC 1
PS 1
Sector 1
3km
Description
LHC PS
FCC PS
FCC MiniArc
Units
Powering sector (PS) length
3000
4000
3200
[m]
Number of PS in the accelerator 8 16 4 -
Filling factor
74%
77%
Number of dipole magnets per PS
154
215
172
Inductance per PS 15
272
218
[H]
Stored energy per PS 1 10
[GJ]
20 PS in total

5. 1 - Powering layout optionsB C D
Option
τdischarge
Energy
EE position
Circuit complexity A B C D E
Example for N=5 E

6. 1 - FCC vs LHC ramp-up
Maximum power to ramp-up dipole magnets in the whole accelerator
Independent of the circuit layout
MiniArcs included
Losses and inefficiency non included
~280 MW
~240 MW
14 MW

7. 1 - Distribution of maximum power flows
tramp = 20 min constant voltage
tramp = 20 min constant voltage + constant power
Ramp-up of FCC block coil magnets only, MiniArcs included, net power (no losses and inefficiency)

8. 1 - Conclusions
Possible directions for the layout of dipole circuits have been identified for the FCC
Subdivision of a 4 km long half-arc in several dipole circuits seems the most feasible solution for proper circuit protection within the required constraints, while at the same time having all converters and EE systems at the access points (using a SC link).
With the subdivision, the ramp-up is not challenging. Trade-off:
Peak power
Ramp-up time (Machine availability)
Decrease the power peak while preserving the same ramp-up time is possible

9. 2 - Are there hard limits for the magnet design?
A string of magnets can always be protected, for any magnet design and given constraints (voltage withstand level, tcirc, …), by adapting the number of circuits, so by subdividing a powering sector in multiple circuits.
Example for N=5
Magnet powering, trade off: Number of circuits Converter voltage rating (VPC) Ramp time (tramp)
Circuit protection, trade off: Number of circuits
tcirc (busbar & diode size & quench prop.)
VFPA,max

10. 2 - Assuming 2x larger stored energy with same current
cable
circuit
ramp time
nr of EE systems (NEE)
VFPA,max
Magnet
protection LM
2x larger
MIIts
larger
VQ,max
Thot
 More copper in cable
Circuit powering
Lcirc
2x larger
VPC
PPC
Awarm-leads
equal
Acurrent-leads
Circuit
protection
Ecirc
2x larger
tcirc
Abusbar
2x larger
Qdiode
EE switch
equal
EE dump
 Increase nr of circuits or number of EE systems
High priority: Minimize stored energy

11. 2x higher current rating
2 - Assuming 2x larger cable (half number of turns) with same stored energy
Magnet protection LM
4x smaller
MIIts
4x larger
VQ,max
smaller
Thot
equal
Same:
circuit
ramp time
nr of EE systems (NEE)
VFPA,max
Circuit powering
Inom
2x larger
Lcirc
4x smaller
VPC
2x smaller
PPC
equal
Awarm-leads
larger
Acurrent-leads
Circuit protection
Ecirc
equal
tcirc
2x smaller
Abusbar
2x larger
Qdiode
EE switch
2x higher current rating
EE dump
Not obvious what is preferable: Low-Inom & High-LM versus High-Inom & Low-LM

12. 2 – Conclusion: how to reduce circuit complexity
Magnet designers should try to minimize the stored energy. Cos- and block designs have clear advantages as compared to the common-coil design.
High voltage withstand levels of the circuits and all its components are needed to reduce the number of circuits.
High-current low-inductance magnets are favourable in terms of quench voltage and number of circuits, which seem to outweigh the drawbacks
I suggest to design several cos- and block dipoles with the usual EuroCirCol requirements, but with Inom=15-25 kA.
The possibility of independent powering of the dipole apertures, each in series with a RQD/F circuit, should be explored from optics point-of-view. This might significantly reduce the number of busbars, power converters, DFB’s, and current leads.

13. Timeline
April 2016
FCC week
November 2016
ECC WP5 review
May 2017 t
1 presentation about circuit protection and magnet protection (CLIQ)
M. Prioli, “Circuit protection aspects, CLIQ simulations and STEAM” (Link)

14. 3 - Outlook for circuit and quench simulations
Lfilt PC
Rfilt
Cfilt
REE M1 M2
M27
M54
M53
M28 Mn
CLIQ
Example: magnets powered with Ncircuits per PS = 4
To understand if this proposal is feasible, circuit simulations are needed
Quenching magnet
CLIQ protection system
Lumped element model of the other magnets in the chain
Other components (e.g. Power Converter (PC), nonlinear switches)
… Quench protection system, controller of Power Converter
CLIQ introduces a coupling between circuit and quench simulations

15. 3 - Field-circuit coupling
Circuit model
Cadence PSpice
Lfilt PC
Rfilt
Cfilt
REE M1 M2
M18
M36
M35
M19 Mn
CCLIQ M L1
ΔU1 R1 L2
ΔU2 R2
φ, R
2D Electro-thermal Field model I

16. 3 - First CLIQ co-simulation Cos-theta v22b
CCLIQ=20 mF, UCLIQ=2 kV − − + + − − + +
Current [A] and resistance [Ω] evolution
Coupling losses [W/m3]

17. 3 - First CLIQ co-simulation Cos-theta v22b
CCLIQ=20 mF, UCLIQ=2 kV
Hot-spot 312 K
Peak voltage to 120ms
Temperature [K]
Voltage to ground [V]

18. 3 - Conclusion
A new tool for circuit and quench protection simulations
First fully consistent CLIQ simulations
Flexibility enables study of complex circuits/magnets without additional effort

19. Timeline
April 2016
FCC week
November 2016
ECC WP5 review
May 2017 t
1 poster with the study of different CLIQ configurations to reduce the voltage to ground after a quench
M. Prioli, “Strategies to reduce the voltage to ground in the FCC main dipole circuits” (Link)

20. 4 - Validation of co-simulation

21. 4 - Results of co-simulation

22. 4 - Conclusion
Co-simulation has proven to be an effective approach for the study of the CLIQ protection system from the circuit and magnet points of view.
Co-simulation of LEDET and PSpice was validated against LEDET monolithic simulation.
The multi-CLIQ strategy is a promising option for the quench protection of the FCC block-coil dipole.

23. Proposal for next activities
Circuit protection
Engage other actors in the circuit layout study
PCs point of view (Benjamin Todd, Valerie Montabonnet ?)
Availability point of view (Andrea ?)
Infrastructure and operation point of view (Volker Mertens ?)
Quadrupole powering (Daniel Schoerling ?)
Magnet protection
Crosscheck of the tools (COMSOL vs. LEDET)
Multi-CLIQ simulations for all magnet designs
Circuit and magnet protection
Co-simulation of the full FCC circuit with CLIQ

25. Circuits layouts for Cos-theta 14% LL
N. of circuits per PS 1 2 3 4 6 8
N. of circuits in long arcs 16 32 48 64 96
128
Magnets per circuit
215
108 72 54 36 27
Inductance per circuit [H]
122 61 41 30 20 15
Stored energy per circuit [GJ]
(<1.6) (>3)
8.1
4.1
2.7
2.0
1.4
1.0
Ramp time [min]
VPC [V] (<500) (>1000)
1139
570
380
285
190
142
VFPA,max [kV]
VFPA,max,fault [kV]
3.3
2.6
2.4
2.3
2.2
τcirc [s] (<150) (>250)
684
342
228
171
114 85
MIITS [A2s]*109 (<10) (>20) 43 22 14 10 7 5
Abusbar [mm2], ΔT=300K
550
390
320
270
220

26. Circuits layouts for Block coil 14% LL
N. of circuits per PS 1 2 3 4 6 8
N. of circuits in long arcs 16 32 48 64 96
128
Magnets per circuit
215
108 72 54 36 27
Inductance per circuit [H]
123 61 41 31 20 15
Stored energy per circuit [GJ]
(<1.6) (>3)
7.8
3.9
2.6
1.9
1.3
1.0
Ramp time [min]
VPC [V] (<500) (>1000)
1124
562
375
281
187
141
VFPA,max [kV]
VFPA,max,fault [kV]
3.3
2.4
2.3
2.2
τcirc [s] (<150) (>250)
674
337
225
169
112 84
MIITS [A2s]*109 (<10) (>20) 14 10 7 5
Abusbar [mm2], ΔT=300K
540
380
310
270
220
190

27. Circuits layouts for Common c. 14% LL
N. of circuits per PS 1 2 3 4 6 8
N. of circuits in long arcs 16 32 48 64 96
128
Magnets per circuit
215
108 72 54 36 27
Inductance per circuit [H] 62 31 21 15 10
Stored energy per circuit [GJ]
(<1.6) (>3)
9.2
4.6
3.1
2.3
1.5
1.1
Ramp time [min] 20
VPC [V] (<500) (>1000)
863
432
288
216
144
VFPA,max [kV]
1.0
VFPA,max,fault [kV]
3.3
2.6
2.4
2.2
τcirc [s] (<150) (>250)
518
259
173
130 86 65
MIITS [A2s]*109 (<10) (>20) 73 37 24 18 12 9
Abusbar [mm2], ΔT=300K
720
510
410
360
290
250

28. Circuits layouts for CCT 14% LL
N. of circuits per PS 1 2 3 4 6 8
N. of circuits in long arcs 16 32 48 64 96
128
Magnets per circuit
215
108 72 54 36 27
Inductance per circuit [H] 55 18 14 9 7
Stored energy per circuit [GJ]
(<1.6) (>3)
9.6
4.8
3.2
2.4
1.6
1.2
Ramp time [min] 20
VPC [V] (<500) (>1000)
819
409
273
205
136
102
VFPA,max [kV]
1.0
VFPA,max,fault [kV]
3.3
2.6
2.3
2.2
τcirc [s] (<150) (>250)
491
246
164
123 82 61
MIITS [A2s]*109 (<10) (>20) 80 40 13 10
Abusbar [mm2], ΔT=300K
750
530
430
370
300
260

29. 1 - Ramp-up of FCC Block coil circuit
Considered case: constant voltage
4 circuits in the 4 km powering sector
54 magnets per circuit
Lcircuit = 31 H
Pcircuit, max = 3 MW
Constant voltage
9 A/s

30. 1 - Distribution of maximum power flows
25 MW
22.5 MW
Ramp-up of FCC block coil magnets only, MiniArcs included, net power (no losses and inefficiency)

31. 1 - Ramp-up of FCC Block coil circuit
Considered case: constant voltage + constant power
4 circuits in the 4 km powering sector
54 magnets per circuit
Lcircuit = 31 H
Pcircuit, max = 1.6 MW
47% reduction
Constant voltage
Constant power
20 A/s

32. 1 - Distribution of maximum power flows
13 MW
11.7 MW
Ramp-up of FCC block coil magnets only, MiniArcs included, net power (no losses and inefficiency)