Transcript Transmission Line Proposals & Constraints 10 September 2001

JET ICRH System Protection Enhancement. CCFW 37
Arc Detection and Generator Controls
November 2004
Tony Walden, Mark Nightingale, Igor Monakhov, T.Blackman, D.Child, M.Graham,
F.Braun (IPP).
Introduction.
• JET sees the installation of a 3dB Splitter system to power two arrays this shutdown, that bring some
new arc-detection challenges.
• The EP antenna has further new challenges for arc detection, especially across the capacitors.
• A recent incident on a JET VTL highlights the weakness of the current JET VSWR trip that voltage
node arcs hardly change the VSWR so go undetected.
• Damage in components due to be reused in the EP system have encouraged some discussion of the
whole philosophy of the existing blanket system protection principally by VSWR.
•This has lead to the intent to review the whole policy of protection on the JET system.
•More specific protection should be safer and allow operation closer to the limit (higher average power
coupled).
1
JET ICRH System Protection Enhancement. CCFW 37
Existing JET ICRH Protection systems.
These consist principally of a VSWR trip, VTL pressure trip, Endstage overcurrent crowbar,
Endstage limiting for transmission-line maximum voltage and some generator limiting controls
such as grid current etc. They have served quite well but there is scope for improvement.
The switches refurbished for EP have shown some damage and would lend themselves to an
optical detector. Other components may suit Microphonic detectors. Suiting detector to
component and setting levels to suit will help both performance and safety. Having some arc
location information should help the Pilot to best decide what operating adjustments may help
or suggest long term uprating.
The VSWR trip is the main arc protection system for the lines and antennas. There are basic
weaknesses that we want to address: • The effect of different impedance on the Endstage tube at the same VSWR.
• It will not detect arcs at voltage node points.
• It does not readily distinguish between arcs and ELMs.
2
JET ICRH System Protection Enhancement. CCFW 37
ICRH Protection systems to investigate.
•A simple VSWR level trip can be caused by a range of impedance and the effect of the voltage
reflected to the Endstage tube varies from relatively benign Anode Dissipation to potentially
disastrous high voltage or oscillation. The electrical length from matching to tube can be set
such that ELMs or voltage anti-node arcs create the former, classically done with trombones
(expensive). The Endstage matching can also achieve this. More sophisticated reflected power
monitoring should allow trip limits to be better tailored to the reactive component and effect on
the Endstage, providing better protection and higher performance.
•Monitoring the speed of change of reflected power characteristics (especially phase) should be
a good arc detector and discriminate against ELMs. This capability is built into the new EP and
3DdB Splitter control systems and will be explored at JET.
•IPP have looked at Wide Band Noise and our first tests of spectrum at JET suggest our machine
may be better suited to this method. There are also options in the manner of detection that we
plan to investigate.
•Microphones in J1T pick up distinctive sounds from SLIMP arcs particularly. Microphones or
strain detecting fibre-optics applied direct should be less prone to outside noise and the detector
could be characterised to the arc sound.
•Arcs were used for transatlantic communication, maybe we can detect them with a crude tranny
connected to a voltage probe?
MARCONI RULES OK?
3
JET ICRH System Protection Enhancement. CCFW 37
Where to investigate ICRH Protection systems.
J1H is in turmoil but when Spinner have finished we have to do some more transmission line
mods and never ones to miss an opportunity we can see some neat twists: • We need to power the test-bed and Spinner will leave C2 connected. OK but C2 will be used in
Ops so we will transfer the test-bed connection to one of A’s Amplifiers as EP won’t be in Torus
this campaign better to use one of its’ non-Ops amps!
• C2 can now be connected to a small rig in the North wall walkway for arc detection studies.
With fast data and switchable to plasma ops, settings need not be disturbed between test-rig and
antenna, allowing powerful comparison if limited testing access. We have enough spares to
make this a ‘free’ option as shown on the next slide.
4
JET ICRH System Protection Enhancement. CCFW 37
Existing C2 layout in J1H
To the test-bed J4
To the torus
short
circuit
pin
C2
OTL U-bend
Stub
outer
DC
break
Test
load
MTL U-link
Trombone
C2 layout proposed for arc studies
J1H north walkway
V
Voltage probe
(arc voltage monitor)
Arc
ignitor
Spare 1.5 m
trombone
To the torus
short
circuit
pin
OTL U-bend
I
V
V
Stub
Test
load
MTL U-link
Voltage and current probes
(OTL and end-stage arc diagnostic)
I
C2
outer
DC
break
Trombone
Voltage and current probes
(MTL and antenna arc diagnostic)
5
JET ICRH System Protection Enhancement. CCFW 37
• The layout allows to use the same set of OTL/MTL probes for arc detection studies both in realistic configuration with
antenna vacuum/plasma loading and in configuration with controlled ignitor-triggered arcs in short-circuited line with variable
arc location
• Stub and two trombones will be used to match the amplifier in case of ignitor configuration; the availability of two trombones
on both sides of the igniter will make it possible to change the location of the arc along the line with respect to the voltage
antinode and to keep the circuit matched at the same time (optionally the line could be terminated by a stub coming from J4)
• In future the line can be re-connected to the test-bed and C2 stub and trombone can be used for amplifier matching during
the tests instead of J4 stub (spare stub + more straightforward matching)
• Safety issues don’t seem unresolvable - operation of the outer DC break and the test-bed switch provide isolation from the
torus earth both on outer and inner conductor. In extreme cases C2 telescopics could be disconnected in the torus hall
• The probes require the fastest possible (250kHz?) data acquisition channels triggered by a trip together with the rest of the
module C FDA (and ideally with the ICE probe) in case of antenna configuration or by the ignitor triggering pulse.
• The necessity to have separate sets of probes on both sides of the stub is explained by the fact that the stub effectively
short-circuits the low frequency components of the arc RF spectrum and the signal levels could differ by orders of magnitude
in case of OTL and MTL arcs
• The probes could be installed next to the U-links on a combination of short (~300mm) straight sections of lines with the
mounting bosses replacing the existing long straight sections. MTL probes have higher priority.
• The arc-ignitor section of line has to be custom-built (?) - two opposite mounting bosses are needed with slightly different
interfaces - one is for the spark plug (4 mounting holes) and another one is for the voltage probe (8 mounting holes)
6
JET ICRH System Protection Enhancement. CCFW 37
Where to investigate ICRH Protection systems.
So far this is just a collection of comments and thoughts from informal discussions at JET. All
of us have pet schemes and other ideas and at JET we are keen to co-ordinate them. One broken
VTL is enough! Series arcs in capacitors and arcs at voltage nodes are beyond JET’s existing
detection capabilities so we must do something ASAP. Only a short campaign to test in plasma
operations.
This hurried assembly of ideas is intended to use this meeting as opportunity to let you know
what we will be starting soon and trigger your comments and suggestions.
Of course volunteers are welcome too. Franz Braun has already put our Dave Child in a better
direction with the Wideband detection as we have started to look at our normal plasma and arc
spectrum. In Restart we hope to pick this up during antenna commissioning when we usually get
a few arcs.
Better still, if you have a system that can be transported to JET and tried, let us know and bring
it along! These systems should mostly be applicable anywhere with a little adaptation and
adjustment so co-operation is good for us all.
7