Download 3.4 Poloidal Field Power Supply Systems for the EAST Steady State

3.4 Poloidal Field Power Supply Systems for the EAST Steady State
Superconducting Tokamak
FU Peng
3.4.1 Introduction
The EAST superconducting tokamak is an advanced steady state experimental device
being built at ASIPP in China from 1998 to around 2004. The device consists of toroidal field
(TF) and poloidal field (PF) superconducting coil systems, a vacuum system, a power supply
(PS) system, a cryogenic system, a data acquisition and plasma control system, a microwave
heat system, and a plasma diagnosis system. Its main parameters are listed in Table 1.
Table 1. EAST tokamak parameters
Major radius R0
Minor radius a
Elongation Kx
1.7m
0.4m
1.6-2
Plasma current
Toroidal field
Pulse length
1.0MA
3.5T
1-1000s
Triangularity dx
0.6-0.8
Volt second
8
10vs
As an electrical device, a tokamak is made up of four main electrical loads: (a) the
toroidal coils, which establish the stabilizing toroidal magnetic field; (b) the central solenoid
coils, which induce the voltage for gas breakdown and build up, maintain and ohmically heat
the plasma; (c) the plasma equilibrium coils, which control the position and the shape of the
plasma; and (d) the additional heating devices, which increase the plasma temperatures and
drive the plasma current. All these loads require alternating-current/direct-current (AC/DC)
power conversion and voltage, current, and power profiles suitable to meet the tokamak
operational requirements.
In the EAST tokamak, the main loads are the PF (including the central solenoid coils)
power supply, the TF power supply, the plasma position control power supply, the LHCD
power supply, the ICRH power supply, the ECRH power supply, the NBI power supply, and
the auxiliary load, including the cryostat system. With the PF power supply system working in
sequential control mode to minimize the reactive power taken from the grid, the active and
reactive power that the EAST system requires are shown in Table 2 and Fig.1.
Table 2. Power required by EAST equipment
Phase
PF (MW)
TF (MW)
Magnetization
Initiation
Slow ramp up
Burn
Ramp down
25
-106
-35
15
-15
0.2
0.2
0.2
0.2
0.2
170
Heating
(MW)
0.8
0.8
0.8
15.8
0.8
Other
(MW)
4
4
4
4
4
Sum
(MW)
30
-101
-30
35
-10
50
Q
40
30
Ip
20
10
P
0
P--Active power (MW)
Q--Reactive power(MVR)
Ip--Plasma current(50kA)
-10
-20
0
5
10
15
TIME( S )
Fig.1 Active and reactive power required by EAST tokamak
All the TF and PF coils in the EAST tokamak are superconducting. Therefore, compared
with a normal tokamak, the power required by the magnets will be considerable smaller,
especially to the TF system: its power decreases from hundreds of megawatts to hundreds of
kilowatts. In EAST, the TF power supply is realized by a 30V/16KA and 12-pulse converter,
which takes the electric power directly from the local grid.
The PF coils in the EAST require a large amount of power during the plasma initiation
and fast ramp up phase, which lasts for about 1s. Such a high pulse electric power, having a
peak value of more than 100MW, cannot be taken directly from the grid because of
unacceptable electric disturbances to the local utility network. To reduce the disturbance
within an allowable limit, two options can be chosen. One is to use an AC flywheel generator
and converter to supply the power required during plasma initiation and ramp up; another is to
use a switch and resistor network to produce the high power required. IPP, Chinese Academy
of Science, now has a 120MVA/400MJ AC flywheel generator, but considering its operating
and maintenance cost, the commutation switch with resistor network is preferred to realize the
plasma initiation and fast ramp up, and the converter is used to achieve the slow plasma ramp
up. Therefore a new 81.5 MVA substation is being built. Two transformers, whose rated
continuous power are respectively 50 MVA and 31.5 MVA, will be installed, they are
supplied by a double bus-bar and distant power connection points 14 km away from the 110
kV power grid in eastern China. The short-circuit capacity of each 110 kV interconnecting
point is 1980 MVA, the short circuit impedance of each transformer between primary and
secondary is 10.5%. A schematic of the power distribution system in the EAST tokamak is
shown in Fig. 2, [1].
Basically, the load of the tokamak consists of AC/DC converters, which are pulsed. Therefore
the load will have a large harmonic content and high reactive pulse power. These factors will
produce a large effect on the power system. Due to excessive repetitive mechanical and
thermal stresses on the local turbo-generators in the grid, a maximum level for the power step
is required. The overall power derivative must also be limited, due to the time response
171
Fig.2 The power distribution system of EAST tokamak
capability of the power-frequency regulation system. Since there is a maximum voltage
excursion guaranteed to customers, a maximum V/V is permitted to the pulse, leading to a
limitation in the reactive power swing, while the maximum active power is limited by the
system power capability at the time of the pulse. Finally, when making substantial use of
rectifiers, such as for a tokamak power supply system, the voltage and current harmonic level
must be minimized. In the EAST tokamak system, a reactive compensation of the 30MVA and
a harmonic filter has been included. And the negative active power is limited to not more than
–5MW by using a suitable operation mode of the PF power supply [2].
3.4.2 Outline of the EAST PF Power Supply
The poloidal field system consists of 14 upper and lower superconducting coils arranged
around the plasma in the EAST tokamak. The configuration is shown in Fig. 3. During the
first few years, EAST will operate in double-null mode, the PF power supply system is
designed according to this operation mode; and in the same time the power supply can also
operate in signal–null operation of tokamak. Therefore the PF coils set is reduced to 6 coil
pairs by suitable series combination. The equivalent inductance parameters are shown in Table
PF9
PF11
PF7
PF5
PF13
PF3
PF1
PF2
PF4
PF14
PF6
PF8
PF12
PF10
Fig.3 Configuration of the PF coils in EAST tokamk
172
3 and the configuration of the PF system and its power supply is shown in Fig.4. Because of
the special characteristics of the superconducting PF coils, their operating requirements are
listed in Table 4.
Table 3. The equivalent inductance parameters of the PF coil pairs
L(mH)
Coil1
Coil2
Coil3
Coil4
Coil5
Coil6
Coil 1
69.06
Coil2
22.58
52.40
Coil3
8.240
19.90
50.90
Coil4
12.30
17.75
32.03
358.3
Coil5
6.020
6.060
6.024
32.99
110.0
Coil 6
3.806
3.644
3.326
16.10
30.82
41.85
Plasma
0.122
0.095
0.065
0.197
0.170
0.119
Plasma
0.004
Fig.4 PF coils and their power supply system in EAST
Table 4. The parameters required by PF coils
PF rated current
15 kA
Maximum voltage between terminal and terminal
10000 V
Maximum voltage to ground
10000 V
Delay time of quench protection
1s
Max. I2dt in quench
5.1×108 A2s
Maximum PF field ramp rate
7 T/s
The power supply system that supplies the CS and PF coils shall provide the following
functions: (1) to provide controlled current in the CS and PF coils for plasma initiation and
plasma current, shape, and position control, (2) to provide a fast discharge for the CS and PF
coils in case of quench of the CS, PF or TF coils, (3) to protect the PF coils against overvoltage and over-current due to abnormal or faulted operation of the power supplies, (4) to
measure the voltage and current in the PF circuits, (5) to provide grounding and ground
173
leakage current sensing in the PF circuits, (6) to isolate or ground the PF coils and the
electrical equipment of the PF power systems for safe access to the areas where the both the
loads and the power supplies are installed by personnel, as required, during a maintenance
period.
In addition, provision shall be made to reverse the direction of the current in all the coils.
The required current rating of the PF circuits has been determined based on (a) the reference
plasma scenario; (b) other PF current distributions due to plasma equilibrium conditions
different from the nominal plasma scenario; (c) the current variations due to control; (d) the
current variations due to plasma disruption. In addition, the design of the PF power supply
must satisfy the performance requirements given in Table 5.
Table 5. PF power supply system performance requirements
Parameter
Value
Nominal cycle period
2800 s
PF magnetization
20 s
PF dwell time to keep magnetization
20 s
Plasma initiation
0.06 s
Plasma current ramp up
4s
Maximum plasma flat top
1000 s
Plasma current ramp down
4s
The PF system consists of 6 coil pairs, with each pair connected to its own power supply
system. The required voltage and current to be applied to the CS and PF coils has been
determined by consideration of three nominal plasma shape (elongated, round and large
volume), along with voltage required for control of the plasma current, position and shape.
The maximum voltage and current required for plasma current initiation and control are given
in Table 6 and Table 7.
Each EAST PF power supply is composed of rectifier transformers, converters which
provide the slow ramp up and control of plasma, thyristor DC circuit breaker (TDCB) which
are used both to initiate the plasma and as the first action for quench protection, and explosive
breakers (EB) that are used as a backup for quench protection. A distributed control system
(DCS) is used for the control and measurement of all PF power supply system. A typical PF
power supply circuit is shown in Fig.5.
Table 6. Maximum PF voltage required for EAST operation
Coil pair
Phase
Magnetization
Current initiation
Other phase
1 (kV)
2 (kV)
3 (kV)
4 (kV)
5 (kV)
6 (kV)
0.25
2.0
0.46
0.27
2.1
0.49
0.24
1.9
0.42
0.6
4.7
0.98
0.15
1.2
0.39
0.1
0.6
0.18
174
Table 7. Maximum PF current and current ramp rate provided by converter in EAST
operation
Coil pair
1
2
3
4
5
6
Rated current (kA)
12
13
12
13
14.2
14.1
Max. ramp rate (kA/s)
18.2
20.3
18.6
9.2
4.9
3.5
Sensor
Fig.5 PF power supply circuit in EAST
3.4.3 The EAST AC/DC Converter System
The EAST PF magnets are supplied by thyristor AC/DC converters, which provide the
current necessary to produce the scenario and to control the plasma shape and position. This
PF converter of the power supply is rated at a total installed power of about 210MVA. The key
issues in the design of the AC/DC conversion system are low cost, high availability and
reliability, along with reactive power reduction.
The 110kV substation is located at about 200m from the PF converter building. A main
transformer, 110 kV/10 kV, 50 MVA continuous rating, supplies the additional heating power
supply, the TF power supply, and PF power supply, at 10kV intermediate voltage. Inside the
PF power supply building, 12 dry rectifier transformers associated with the converters and
other switches are installed. The typical circuit diagram for the PF power supply converter is
shown in Fig.6.
Fig.6 Simplified diagram of PF converter
175
Each converter operates in four quadrants with circulating current between the head and
tail sets through the inductance L1, L2, L3 in series. Converter G1 consists of two half bridges
G11and G12 in parallel, G11 and G12’s angle difference is 180 degrees, and they share one phase
angle controller. As do the converters G2, G3, and G4. For G1 and G2, G3 and G4 upper and
lower, each set head or tail is made of two series identical Graetz bridges with a 30 degree
angle difference between the two transformer outputs, and the 12-pulse output voltage. G1 and
G2, G3 and G4, which are respectively paralleled for one tail and head set by inductance L1 and
L2, make use of the same transformer secondary. The operating principle is based on the
simultaneous conduction of only two of the four bridges selected automatically according to
the amplitude and the direction of the total load current. An integrated cooling water
exchanger extracts the heat energy produced in the bridges.
When the coil current is in the range from 10% to 100% of the maximum operating
current, the main AC/DC converters in the PF coil circuit operate in two quadrants mode, in
which G1, G3 drive the positive load current and G2, G4 drive the negative load current.
When the coil current is below the 10% level, the converters operate in the four-quadrant
mode, in which only G11, G31, G22, and G42 operate. This operation mode of the PF
converters can save half of the capacity of the required transformer, and also ensure that the
load current crosses zero smoothly from positive current direction to negative current
direction.
The synchronous operation of the two series power bridges of each converter permits a
twelve-pulse waveform on the AC network. At low load voltage, the reactive power
consumption is minimized in the converters by a firing offset between the two series power
bridges. In this case, the twelve- pulse reaction can approximately decrease the maximum
reactive power 50 MVAR to 25 MVR in the PF power supply system of EAST [3].
3.4.4 Direct-current Circuit Breaker
The function of the DC circuit breaker is to insert, in the tokamak coil circuit, discharge
resistors, which are used to absorb the stored magnetic energy from the coils. In each PF
power supply of EAST, DC breakers are used during the plasma initiation and current rampup, its parameters are shown in Table 8.
Table 8. The rated parameters of TDCB in EAST
Rated current
15 kA
Rated voltage
2.4 kV
Rated switching time
2 ms
Dwell time before repeatable switching
200 ms
Energy dissipated in plasma initiation
8 MJ
Energy dissipated in quench protection
20 MJ
176
Since this DC circuit breaker system is subjected to very frequent operation, both the
electrical and mechanical life of the system are of prime importance. In the plasma initiation
phase, each PF circuit must contribute initiating plasma. Therefore each of the DC breakers in
each power supply must be well synchronized. In the EAST power supply, the required
current and voltage parameters of the DC circuit breaker are not extremely high, therefore a
TDCB associated with resistor network is chosen, which can also switch bi-directional current
and act as the main switch for quench protection. The TDCB consists of thyristor, diode,
capacitor, resistor and inductor. Its parameters are given in Table 7,and the circuit is shown in
Fig.7. The two thyristor sets and two diode sets (Th1, Th2, D1, and D2) permit bi-directional
current and voltage.
Fig.7 the TDCB circuit diagram
The fast discharge circuit is composed of capacitor C1, L1, th3 and th4, C1 is pre-charged
to 2.4 kV. When the positive current of power supply flows through Th1 and D1, while Th3 is
triggered to discharge the capacitor C1 through the series circuit including Th1, D1 and L1.
The discharge produces an artificial current zero of current in Th1, which turns off Th1. The
full current of the power supply is transferred into the C1 branch. When C1 is recharged in the
opposite direction, the current is progressively transferred into the discharge resistor connected
in parallel with the switch. When the opposite current flows through Th2 and D2, the
commutation is realized by triggering Th4.
During the commutation, the resistor network produces a high voltage to initiate the
plasma discharge, and dissipate the energy stored in superconducting coil in quench
protection. The resistor is made from stainless steel tube in a tight serpentine pattern to
177
minimize the inductance, and it is cooled by water. Each resistor unit consists of several
modules connected in parallel or in series. This arrangement allows the resistor value to be
adjusted according to the requirement.
A RC snubber circuit is employed to absorb the instantaneous over-voltage during TDCB
operation. When TDCB opens a small current load, it makes the coil current to increase,
because of excess energy in capacitor C1. Therefore D1 and D2 are respectively paralleled
with Th1 and Th2. A dynamic and static current distribution in the parallel thyrisor is also
ensured by careful layout of the structure.
Quench protection for the superconducting magnets is of critical importance because of
the large amount of stored energy. In each power supply, one TDCS is used as the main
quenching protection, and two explosive breakers (EB) are as stand-by switches.
3.4.5 PF Power Supply Control System
The local control system of PF power supply provides the routine control, real-time
control, communication of control data, a timing system for precise control of event initiation,
and data acquisition. The PF power supply system has more than 140 analog signals and 600
digital signals, which will be measured, supervised and controlled. A distributive control
system (DCS) consisting of industrial personal computer (IPC) computers has been used, and
all computers are connected as real time Ethernet network by the QNX real time operation
system. In addition, a field-bus network is also used on the spot. The overall structure of the
PF power supply control system is shown in Fig.8.
NET:TCP/IP INTERNET
System State
Display
Fault
Diagnosis
Supervisory
Control System
Waveform
Analysis
Center Contol system
PF Current
Feedback controller
QNX Server
NET:FLEET
AD,DI&
Database
CAN
Controller
PF1 subsystem
controller
PF2 subsystem
controller
PF6 subsystem
controller
NET:CANBUS
AC
components
PF1
components
178
PF2
components
PF6
components
QNX is a real time operating system, developed by QNX Software System Limited
(QSSL) in Canada, and is widely used in industrial PC’s. Besides its high performance in real
time, reliability, and embedded characteristics, the unique feature of the QNX real time
operating system is its network technology. Its FLEET is an ultra-light, high-speed networking
protocol. Its innovative and feature-rich design turns isolated machines into a single logical
supercomputer. Because FLEET is built on the message passing architecture of the QNX OS,
it offers the very high flexibility. Its features are fault-tolerant networking, load-balancing on
the fly, efficient performance, extensible architecture, and transparent distributed processing.
Via the QNX all IPC computers in PF power supply control system can be connected as a
single large computer, and all IPC share the data and signal in a high speed and transparent
network.
The distance between the power- supply building and its control room is 40m. To
decrease the transmitting cable and disturbance, A network consisted of CANBUS has been
built. Many CAN modules work at the location of the power supply. All supervised signals
including a majority of control and protection signals, and a majority of analog signals are
transmitted by CAN BUS. Other critical signals required for high real time performance are
Fig.8 Overall structure of the PF power supply control system
transmitted directly by wires.
All logical signals and A/D conversion of analog signals measured by the IPC and fieldbus modules is transmitted to a database computer [4].
3.4.6 PF power supply R&D
Simulation studies by Simplorer software and laboratory tests of PF power supply in
EAST have been completed. They show that the design of the PF power supply system is
feasible and reliable. All of the equipment of the PF power supply is being fabricated by
industry. Now one set of PF power supply has arrived at the institute and been installed [5].
Reference
[1] E.Bertolini, et all., The JET magnet power supplies and plasma control systems.
Fusion Technology. vol.11, pp71-119, Jan.1987.
[2] R.Shimada, et all, JT-60 power supplies. Fusion Engineering and Design. pp47-68,
May, 1987
[3] Benfatto I. et al., AC/DC Converters for the ITER poloidal system in Proc. of the 16th
SOFE, 1995. Champaign, IL, USA, pp658-661.
[4] C.Sihler, P.Fu, M.Huart, B.streibl and W.Treutterer. Paralleling Of two large
Flywheel Generation for the Optimization Of the ASDEX upgrade Power Supply. 21st
Symposium on Fusion Technology, Madrid, Spain, Oct., 2000.
[5] SIMEC GmbH. Simplorer. Version 4.2, Chemnitz, 2000.
179