Association
Euratom-Cea
TORE SUPRA
The control of magnetically confined plasmas
(in tokamak facilities)
S. Brémond
Institut de Recherches sur la Fusion par confinement Magnétique (IRFM)
CEA Centre de Cadarache
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 1
Outline
Association
Euratom-Cea
TORE SUPRA
• A short reminder of the fusion energy source development issue
• Basics of tokamak operation
• Control issues overview
- Basic controls
- Performance optimisation, advanced scenario
- Machine protection
• Control design needs
• Conclusion
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 2
A SHORT REMINDER OF THE FUSION
Fusion reaction ENERGY SOURCE DEVELOPMENT ISSUE 1/3
Association
Euratom-Cea 9 Fe
Binding energy by nucleon (MeV / nucleon)
8
Ne
TORE SUPRA O
C U
7 He
FISSION of heavy nucleus
6 Be FUSION of light nucleus
D + T He + n
5 Li
(3.6 MeV) + (14 MeV)
4
3
T
2
Requires to give the two light nucleus enough
energy so that they have a chance to jump through
1 D
the electrostatic barrier Nb of
0
nucleons
0 50 100 150 200 250
FUEL AND HEAT IT UP ! (to about 100 million degrees, plasma state)
EXTERNAL HEATING required
before internal heating takes over (collisions of the He with the D-T nucleus)
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 3
A SHORT REMINDER OF THE FUSION
Power balance ENERGY SOURCE DEVELOPMENT ISSUE 2/3
Association
Euratom-Cea CONFINE IT (to avoid power losses and fuel dilution) ! Gravitational confinement
Condition for overall power gain (Lawson criteria)
TORE SUPRA
Huge mass
n x T x tE > threshold required
(sun and stars)
Fuel density Temperature Confinement time
G
Inertial confinement
G 1
Compression of a
milimetric target
n ~ 103 x standard solids
1995- tE ~ 10-9 seconde
JET
1990
Magnetic confinement
1980
Effect of magnetic field
Pfusion on charged particles
1968 G
Pexternal heating
n ~ 10-5 air in this room
tE ~ 1 seconde
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 4
A SHORT REMINDER OF THE FUSION
Magnetic confinement 1/2 ENERGY SOURCE DEVELOPMENT ISSUE 3/3
Association
Euratom-Cea
The Issue: losses at
principle both ends
TORE SUPRA
Issue: vertical drift
B = m0 Ibob / 2p R
The toroidal
configuration Giration radius a 1/B
Issue (collective
Helicoïdal effects between
I particles) : hoop
field lines p force
The tokamak configuration
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 5
A SHORT REMINDER OF THE FUSION
Magnetic confinement 2/2 ENERGY SOURCE DEVELOPMENT ISSUE 3/3
Association
Euratom-Cea
Equilibrium
TORE SUPRA
field
external
coils
SET OF MAGNETIC
CONFINEMENT COILS
Two options
Wall
Scrape-off layer (SOL)
Last closed flux surface
Limiter configuration divertor configuration
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 6
BASICS OF TOKAMAK
Tokamak plasma actuators 1/2 OPERATION 1/4
Association
Euratom-Cea
Central
TORE SUPRA solenoid coil
(transformer
FUEL primary)
Gaz valves, pellets injector,
Pumps
Poloidal
CONFINE Field coils
Fuelling (plasma
-Transformer primary coil system (gas position
(limited time duration) + non valve, pellet and shape)
inductive current drive systems injector not
-> drive plasma current represented
here)
- Poloidal Field coils
-> set plasma position and
shape Pumping
(cryo
pumps)
HEAT UP
- “ohmic” heating (limited
capabilities)
- Wave, Beam heating systems ITER project view
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 7
BASICS OF TOKAMAK
Tokamak plasma actuators 2/2 OPERATION 2/4
Association
Euratom-Cea
Heating / Current Drive
TORE SUPRA
NEUTRAL BEAM
ION CYCLOTRON WAVE
Tens of MHz (tetrode sources)
ELECTRON CYCLOTRON
WAVE
Around 100 GHz (gyrotrons)
LOWER HYBRID WAVES
2.5 GHz, 3.7 GHz (klystrons)
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 8
BASICS OF TOKAMAK
Example of Tore Supra (CEA Cadarache) OPERATION 3/4
Association
Euratom-Cea
PERMANENT TOROIDAL MAGNETIC FIELD :
supercondutctor NbTi coils
TORE SUPRA
ACTIVELY COOLED PLASMA FACING
COMPONENTS : water cooling, exhaust
capability ~ 10 MW/m2 (within ITER range)
Circular cross-section
Current 1.5 MA
CONFINED
Major radius 2.4 m
PLASMA
Minor radius ~ 0.72 m
Volume ~25 m3
2 Lower hybrid 3 Ion Cyclotron 1 Electron Cyclotron
antennas (3.7 GHz) antennas (40-80 MHz) antenna (118 GHz)
~ 6 MW max, 1 GJ ~ 10 MW max, 250 MJ 0.8 MW max , 25 MJ
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 9
BASICS OF TOKAMAK
Plasma discharge schedule OPERATION 4/4
Association
Euratom-Cea ITER discharge foresseen schedule
TORE SUPRA
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 10
CONTROL ISSUES 1/2
Control issues overview
Association
Euratom-Cea second
-3 -2 -1 0 1 2 3 4
TORE SUPRA 10 10 10 10 10 10 10 10
Plasma MHD Energy confinement Current diffusion
Thermal ELM’s - Disruption Cooling of PFCs
Particles Particle confinement Wall inventory
Basic controls Machine protection
(MHD equilibrium, density) Safety
Performance optimisation
Advanced scenario
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 11
CONTROL ISSUES 2/2
Feedback control need
Association
Euratom-Cea Target scenario
CONTROLER / DIAGNOSTICS
TORE SUPRA
• INTRINSIC INSTABILITIES SUPERVISOR measurements
• PERTURBATIONS (state of the Sequencing, tracking of
references and machine
wall, internal profiles “self- protection
organisation”, etc. )
• COMPLEXITY OF THE PHYSICS at
stake (accurate prediction very ACTUATORS Tokamak
plasma
difficult) - magnetic coils
- fuelling / pumping systems
- heating / current drive
systems
Real time feedback control required
Reliable real-time
measurements
needed
(not covered here)
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 12
JET diagnostics overview
BASIC CONTROL 1/4
Equilibrium control – basics 1/3
Association
Euratom-Cea BASIC MODELING (PLASMA POSITION)
TORE SUPRA
MOTION EQUATION
(Magneto Hydro Dynamic
dv
dt
v
t
v . grad v j B grad p
description)
RIGID DISPLACEMENT MODEL r ( t )u r z ( t )u z
m p R 2pf R m p Z 2pf Z
f jt B p plasma jt B p ext j p Bt grad p
m0 I p
2
Zero order m0 I p
2
m0 I p p 1 8p u R p
2
development 8R li RI p BZext u R BRe xt u Z uR
ln 1u R 8p
in R 4p a 2
a
m 0 I p 8R
2
l 3
m p R 2pRI p BZext ln p i
m p Z 2pRI p BRe xt
2 a 2 2
(Poloidal) beta: ratio of
kinetic / magnetic pressure Plasma self inductance
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 13
BASIC CONTROL 2/4
Equilibrium control – basics 2/3
Association
Euratom-Cea VERTICALLY ELONGATED CROSS SECTION
TORE SUPRA - happens to allow better plasma performance
require outward curvature
- but require outward field curvature
Vertical position unstable
Huge force : Tesla * MAmp * 2pR
Low mass: 800m3 * 1020 (per m3) * 10-27 (kg/ion) • •
Very fast time scale : ~ ms (MHD Alfven time)
But we don’t have to react that fast (we could not actually)
Ip
because eddy current will be induced in the metallic
structures of the vessel which will – on certain conditions-
slow down the motion to their L/R characteristic time where
feedback controlled active coils may take over
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 14
BASIC CONTROL 3/4
Equilibrium control – basics 3/3
Association
Euratom-Cea
PLASMA SHAPING
TORE SUPRA
ITER Poloidal cross section
Numerical Modeling :
•Direct / Inverse problem (PF coils current -> magnetic surfaces topology or the inverse): 2D
free boundary equilibrium codes (CEDRES++ developed with Univ. Nice)
•Reconstruction problem (magnetic probes at the vessel -> plasma last closed surface): 2D
free boundary reconstruction code (EQUINOX developed with Univ. Nice, other approach for
plasma boundary reconstruction only developed with INRIA Sophia)
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 15
BASIC CONTROL 4/4
Particle density control
Association
Euratom-Cea
ne
TORE SUPRA FUELLING SOURCES
Gaz
•Recycling : a badly controlled dominant puffing
fuelling source (depending of the history / +
state of the wall) Recycling
•Gaz puffing : edge fuelling
•Pellets injection : more internal fuelling Pellets
Complex particle transport phenomena
Control target / issues: Gaz puffing
pellet
- impurity exhaust (He ashes)
- density profile optimisation (burn control)
- radiation control (divertor) recycling
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 16
PERFORMANCE OPIMISATION –
Advanced scenario ADVANCED SCENARIO 1/2
Association
Euratom-Cea
• internal confinement enhancement
Control of the internal profiles (plasma
TORE SUPRA
current density, pressure, etc.)
• MHD instables modes control
Edge localised modes, sawteeth,
neoclassical tearing modes, resistive
wall modes, etc.
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 17
PERFORMANCE OPIMISATION –
Internal profiles control (current density) ADVANCED SCENARIO 2/2
Association
Euratom-Cea
Distributed plasma current
density real-time
TORE SUPRA
measurement
RT equilibrium reconstruction with the EQUINOX
code (developed with Univ. Nice)
Identification of the plasma current profile
(and free boundary equilibrium) from external
measurements
Distributed plasma current density
real-time control design
New design approach based on a distributed non linear control oriented model
// jT
1
x // R 0 j ni (1, t ) x
m 0 R 0 a 2 x x x
t m 0 a 2 x x x
a 2 xB 0
q
“Diffusivity - interior - boundary” control x
magnetic flux, //parallell electric resistivity, jninon inductive current density
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 18
MACHINE PROTECTION ISSUES 1/1
Machine protection issues
Association
Euratom-Cea
• Disruption detection
TORE SUPRA
• Disruption mitigation
• Plasma First Wall Component protection from overheating
Needs :
- IR imaging wrapping to PFCs geometry
- High level real time data processing
(pattern recognition, heat flux
assessment,multi diagnostic processing,
etc.)
collaboration under way with INRIA Sophia
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 19
CONTROL DESIGN PROCESS 1/1
Control design process
Association
Euratom-Cea
• Existing control mainly based on “semi empirical” design
TORE SUPRA • Model based control: needs for integrated modeling
Plateforme de
simulation
(Kepler)
Boite à outils
contrôleurs Modèles orientés
(Scicos) contrôle
Paramétrage
avec éditeur TS
TS
Structure de
données ITM Contrôleurs
Paramétrage avec éditeur
générique ITM Diagnostics TR
Installation actuelle Outils ITM en cours Briques ou connexions
Tore Supra de développement à développer
Plasma discharge flight simulator under development
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 20
Conclusion
Association
Euratom-Cea
TORE SUPRA Control of tokamak plasmas: an old problem, new issues
Performance target requires to (ITER needs) :
- operate very close to technological limits (while ensuring machine
protection)
- optimise plasma performance by controlling not only global variables,
but also local ones (plasma internal profiles)
- develop an integrated management of the plasma discharge
-pre-qualify control algorithms (safety – licensing)
Main needs
- real-time data processing algorithms development
- Integrated modeling tools (model based control design)
Large Scale Initiative « fusion » Summer School (September 2009) S. Brémond 21