Transcript Euratom

Euratom
TORE SUPRA
Arc Detection in ITER-like structures
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Introduction
Voltage breakdown in IC arrays, a frequent occurrence in today tokamaks, is dealt with by
a quick, temporary suppression of the power flow, and power application after the arc
extinction.
The primary goal is to extinguish the arc and to limit the amount of energy deposited on the
surfaces to avoid permanent damage, as a local damage often becomes the reason of further
arcs in the same region..
In ITER the damage must be limited to the extent that cooling channels, located few mm below
the surfaces are not affected, as a coolant leak may have serious consequences on the safety
of the operation of the torus and should be prevented.
On the other hand, as the extinction of the arc requires a temporary suppression of power,
a false detection of an arc is as well undesirable, because it reduces the the overall
efficiency of the heating system.
An abrupt power suppression in undesirable also because it tend to destabilise the
overall control of the array, which relies on various feed back loops (power, phase,
impedance match…) which need to be orderly restarted when the power is re-applied, to
avoid inducing mismatch conditions, which is can be misinterpreted by the protection
systems as the condition of further breakdown.
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Introduction
The proposed ITER IC launcher is relatively
complex array of 24 straps, in CEA
proposal combined in pairs in individually
powered structures
In IC arrays fed by multiple sources, the
occurrence of a power trip in one of the
elements tends to affect the matching of
other elements and very often the whole
array is shut down
In ITER RF power, trips in the whole
launcher are particularly undesirable,
because in a driven operation, the plasma
energy content is supported by the heating
power and an abrupt loss of this, may
cause a b limit and a disruption
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Introduction
Present arc detectors are in general triggered by an excessive value ofthe VSWR
measurements, performed at the generator end, outside the resonant matching loop.
The effectiveness of this method relies on the the fact that, in general, perfect match is
rather critically load dependent, and fast load changes due to a breakdowns are reflected in
VSWR changes.
Automatic matching systems reduce or eliminate this criticity and may end up matching an
arc. This is particularly undesirable, because the arc is not only undetected, but a large RF
power is injected in the fault and can cause severe damage. Such an accident has caused in
JET the break of a ceramic feed through and an unscheduled shut down.
Load variations too fast to be tracked by the automatic tuning system, such as those caused
by ELMs, have effects similar to break-down and cause false trip requests.
“Load resilient” structures on one hand may help reducing the number of false trip
requests, but the inefficiency of this arc detection method is substantially increased.
The operation of the Tore Supra ITER Prototype has experimentally shown arc detection
based on VSWR measurements may fail detecting voltage breakdown in ITER like stuctures.
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Arc damages in TS ITER prototype
Area of arc damage
The damage occurred in a low voltage area (V < 5
kV peak) due to the absence of a corona ring
which should have been installed and was not.
The damage probably developed during high
power test in vacuum
The arc protection system (at generator end) used
in standard TS antennas and protecting the power
source was in operation throughout the test
Light emission was detected the antenna front,
but interpreted as the result of multipactor
A set of four voltage probes is instakked in the TS
prototype, which were put in operation on plasm
Position of voltage probes
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Results of arc damage on TS Proto
Arc detection based on probes in the resonant part
of the circuit may be crucial to avoid these events
Observed damages likely to have been produced
by unsuppressed arcs lasting beyond the response
time constant (about 100 ms) of the system
protection
Traces of arcs and marks of burn in the bridge area
and complete mechanical detachment of the
brazing of the capacitors (understood as the
consequence
of the omission of a corona
ring at
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
the capacitor low voltage side).
Euratom
TORE SUPRA
E-field enhancement in the brased joint
Dielectric
Torus ’vacuum’
Private vacuum
Copper
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Current Probes for Array Control and Protection
The control method we have proposed for the
ITER array (1) relies on the vectorial control of all
array total currents.
G. Bosia et al. Final Report EFDA task 1129 (2005)
It can be shown by FEM analysis that the local
voltage or B-flux measurements by probes
performed in the antenna box can be strongly
dependent on plasma loading and local conductivity.
If this is actually true these measurements are not
representative of the total current
We are therefore investigating methods for
detecting the total current in some sections of
the antenna circuit in both the TS proto and
ITER
It is advisable to include the design of the
detectors in the overall array design. This we
have not done for TS Proto...., but it can be
done for ITER.
G. Bosia, “Arc detection in ITER-like structures”
Position of the voltage probes in
TS
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Possible probe positions in TS Proto
Position 2
Current probe
Position 1
Vectorial impedance
probe
2 Current + 1 voltage
probes
Position of standard voltage probes in TS ITER Proto
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Vectorial (Voltage + Current probe) at the T-junction input
This is the « impedance » probe proposed for ITER in 1)..
1) G. Bosia et al. Final Report EFDA task 1129 (2005).
Vectorial prob
V0
I1
I2
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
A Vectorial Monitoring Device
The impedance monitor, is located in the symmetry point of the T- junction and
consists of a voltage probe in series with two current loops, each monitoring the
currents I1 and I2 in the two half sections. The three outputs of the probe (V, V1 and V2 )
are connected by identical coaxial cables to signal conditioning networks, located
outside the vacuum vessel.
C1
V0
I1
I2
V1
V  i    k  I1
V2
V  i    k  I2
C2 

V  1  2 
  i    k  I1  I2
C1


V0

Impedance probe layout

V
I*
V1
k*
L
IV
10
R
I*
k*
LI2
R
C2
V2
C2
Equivalent circuit (including signal trasmission)
From the three output, any ILS input vectorial quantity (impedance, admittance, reflection
coefficient..) can be deduced by simple linear combinations
G. Bosia, “Arc detection in ITER-like structures”
RL
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Current monitor for TS ITER Proto
If the circuit is far from resonance,
L>> Rc+Rl and close coupling
d

dt
Isolating material.
Current transformer
Current Strap
Ground potential
the coil acts as a current transformer
and the measurement is essentially
independent of frequency:
i
d

dt
RL
G. Bosia, “Arc detection in ITER-like structures”

L
Lc Rc
d
L i
dt
m 0 NA  l
S
1

2
S
m 0 N A  l N
Practical difficulties :
1) capacitance to ground to be eliminated
by a balanced transmission
2) It is difficult to design a high Q, high
frequency transformer in air at 60 MHz.
3)The probe can be used in TS and JET
but is unlikely to be used on ITER
V
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Views of the current monitors
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Proposed scheme for control of the RF currents in a
generic ITER-like structure
Main Transmission
Line
I1
I2
V
Decoupler
Trimmer
Pretuner
Isolator
ILS control
Source end
Load end
G. Bosia, “Arc detection in ITER-like structures”
Decoupler
CCFW 39 ,JET, 4th August 2005
RF source
Euratom
TORE SUPRA
Compact tuner
Ground
& shield
Straps
Feeders
G. Bosia, “Arc detection in ITER-like structures”
Ceramic septa
CCFW 39 ,JET, 4th August 2005
Capacitive tuners
VTL
Euratom
TORE SUPRA
Possible probe positions in compact tuner
Position A : T- junction input
Position B : Strap Short circuit
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Position C : T- Capacitors SC
Euratom
TORE SUPRA
Coaxial Shunt detector
In coaxial shunt detector the condition R >>  L is obtained by shielding the TEM field
of the main current path from the sensing probe . The voltage drop along the resistive
cilinder is sensed by the connection made at the end with respect to ground.
•
Ez = J0/s exp[x/d(1+ i)]
For SS or inconel , d ~ 0.1mm, which is the
thickness of a bellows.
I
R
V= R I
•
A simple, robust layout could results, with a wideband vectorial measurement of the
RF current. As the resistor thickness d ~ d, the resistance R requires calibration vs
frequency.
•
The coaxial probe dissipates a significant amount of RF power. For I = 103 A and
R = 10 mW Voutside = 10 V outside and P = 5 kW i.e. The shunt requires water
cooling. This is not a limiting factor in iITER, where everything is water cooled
•
This also means that the probe is in practice an ideal voltage generator that can be
used to directly drive in-vessel opto-electronic devices
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Use of tuners bellows as current probes
The (cooled) bellows of a coaxial
capacitor can be in principle be used
as a shunt current monitor, if the
driving shaft is electrically isolated
from ground potential.
I
R
“ground” potential
Driving shaft
V
V= R I
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
General layout for current monitors in ITER
Cooled coaxial
structure in the
system under
private vacuum
Flexible isolated signal transmission
line (matched at both ends)
Fast LED
Or exit signal line
at “service stub”
using an Isolating
transformer
Cooled resistive section,
floating at some high RF
potential
G. Bosia, “Arc detection in ITER-like structures”
IR receiver at
ground potential in
a neutron shielded
position
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Current
straps
Arcs in “Standard” TS antennas
30 W feed line
Port
flange
Vacuum window
Capacitors housing
l/4 service stub
(cooling)
Capacitor positioning unit
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Voltage
Power reference
Maximum plasma
operation RF voltage
(ends of straps)
Loading resistance
Reflected power
60 kV 500 ms
45 kV currently
achieved
Plasma position
Plasma density
Plasma current
G. Bosia, “Arc detection in ITER-like structures”
Tore Supra pulse 28119
CCFW 39 ,JET, 4th August 2005
Versatile
pellet injection system for fuelling studies
Euratom
TORE SUPRA
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Top view LFS
Pellet injection system relevant for ITER
Euratom
TORE SUPRA
• Control of the density by pellet injection
– Modification of the edge density
• The antennas far away from the injector are not
perturbed
• Each pellet leads to a power cut of the closest
antenna
• Notching of LH wave allows pellet penetration
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
2 min. pellet injection with LH
0.6
MW, 10 m
-2
Euratom
0.4
Vloop (V)
0.2
2
<ne>
1
<M> = 1.5 1020 atoms
0
0.1
LH power (MW)
2
1
0
0
20
40
60
80
Time (s)
G. Bosia, “Arc detection in ITER-like structures”
100
nl reference
97
t (s)
98
99
LH power notching allows
penetration of 155 pellets
Very stable speed of 0.5 km/s
V = ~ 0.5km/s
0.0
3
PLH
2.5
96
(1019m-3)
nl(0)
19
Ip (MA)
3.0
TORE SUPRA
120
CCFW 39 ,JET, 4th August 2005
Euratom
Pellet
effect on nearby ICRF antenna
Central line density
Ref.
TORE SUPRA
#33952
Ptot = 10 MW
ICRF Injected power
Ant. 1
Ant. 3
Ant. 2
LH Injected power
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
fGreenwald = 0.8
The antenna tripping is
the closest to the pellet
injector
•Simulation of ELM
effect
•Test Bed for the ITER
like antenna
Euratom
Résultats
TORE SUPRA
densité & chauffages
10
Choc 34899
8
6
4
5.6s au dessus de Greenwald
2
0
0
2
4
6
8
10
• Très fortes densités obtenues sans être confronté à la limite de Greenwald ( 180% de nG
obtenu sur le choc 34899 ) en LFS
• La phase au dessus de Greenwald stable de 5,6s sans problème.
• Consigne en densité très bien suivie. Elle n’a été arrêtée que sur des coupures FCI (
limite radiative)
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Effect of non suppressed external arc on the high voltage side
Q1H
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005
Euratom
TORE SUPRA
Effect of a non suppressed internal arc
Q2H
G. Bosia, “Arc detection in ITER-like structures”
CCFW 39 ,JET, 4th August 2005