Plasma Science and Technology, Vol.16, No.12, Dec. 2014
On the Sequential Control of ITER Poloidal Field Converters for
Reactive Power Reduction∗
YUAN Hongwen (袁红文)1 , FU Peng (傅鹏)1 , GAO Ge (高格)1 ,
HUANG Liansheng (黄连生)1 , SONG Zhiquan (宋执权)1 , HE Shiying (何诗英)1 ,
WU Yanan (吴亚楠)1 , DONG Lin (董琳)2 , WANG Min (王敏)2 ,
FANG Tongzhen (房同珍)2
1
2
Institute of Plasma Physics, Chinese Academy of Sciences, Hefei 230031, China
China International Nuclear Fusion Energy Program Execution Center, Beijing 100862,
China
Abstract
Sequential control applied to the International Thermonuclear Experimental Reactor (ITER) poloidal field converter system for the purpose of reactive power reduction is the
subject of this investigation. Due to the inherent characteristics of thyristor-based phase-controlled
converter, the poloidal field converter system consumes a huge amount of reactive power from the
grid, which subsequently results in a voltage drop at the 66 kV busbar if no measure is taken. The
installation of a static var compensator rated for 750 MVar at the 66 kV busbar is an essential
way to compensate reactive power to the grid, which is the most effective measure to solve the
problem. However, sequential control of the multi-series converters provides an additional method
to improve the natural power factor and thus alleviate the pressure of reactive power demand of
the converter system without any additional cost. In the present paper, by comparing with the
symmetrical control technique, the advantage of sequential control in reactive power consumption
is highlighted. Simulation results based on SIMULINK are found in agreement with the theoretical
analysis.
Keywords: sequential control, power diagram, poloidal field converter, ITER
PACS: 84.70.+p
DOI: 10.1088/1009-0630/16/12/11
(Some figures may appear in colour only in the online journal)
1
Introduction
voltage output in this period. To solve this problem,
a set of static var compensator (STATCOM) [6] rated
for 750 MVA is installed at the 66 kV busbar to provide the required reactive power to the grid. Sequential
control [7−10] of the multi-series converters is also an effective method to improve the natural power factor of
ITER PF converters thus to alleviate the pressure of
reactive power demand from the grid. It plays a supplementary role in improving the performance of the
converter system without adding any cost. This paper
mainly describes the principle of the sequential control
strategy for more than one converter module in a series. Simulations based on SIMULINK are performed
to verify the theoretical analysis.
International Thermonuclear Experimental Reactor
(ITER) [1] is a scientific and engineering project under
construction in the south of France in order to demonstrate the feasibility of fusion energy for peaceful uses.
ITER poloidal field (PF) converter system [2−4] , as a
subsystem of the ITER machine, is of significant importance for plasma control in vertical and horizontal
directions. There are six PF superconducting coils embedded in the ITER machine that are fed by 16 sets
of converter modules, each of which is designed to produce DC voltage and DC current rated for ±1.05 kV
/±55 kA. The requirement and assembly pattern are
shown in Table 1, in which it is indicated that PF1 and
PF6 are respectively supplied by two converter modules in a series, and PF2-PF5 respectively by three in
a series.
Due to the inherent characteristic of thyristor
phase-controlled converters, a huge amount of reactive
power [5] is required by the PF converter system during normal operation, especially at the flap top phase
of plasma current because of the high current and low
∗ supported
Table 1.
Coils
PF1
PF2
PF3
PF4
PF5
PF6
Vground =12 kV
ITER PF circuit requirement
Rated voltage
(kV)
±2.10
±3.15
±3.15
±3.15
±3.15
±2.10
Number of converter
units in series
2
3
3
3
3
2
Idc = 55 kA
by International Cooperation Project of Ministry of Science and Technology of China (4.1.P2.CN.01/1A)
1147
Plasma Science and Technology, Vol.16, No.12, Dec. 2014
2
2.1
Principle of sequential control
Two converters in a series
The series connection of two six-pulse converter modules is shown in Fig. 1. Vtotal is the total reference
voltage obtained from the master controller. V1 , V2
represents the reference voltage for converter modules
#1 and #2, respectively. Ud1 and Ud2 are the output voltages of the converter modules; Ud is the total output voltage. Each converter is supplied by one
single transformer. Fig. 2 illustrates the principle of
the symmetrical and sequential control method for two
six-pulse converters in a series upon the reactive power
consumption.
Fig.1
Two converter modules in series
Consider ideal conditions by ignoring the commutation overlap angle brought by leakage inductance of
converter transformer. In the symmetrical case, to output the required dc voltage, each six-pulse converter
shares half of the total reference voltage.
V1 = V2 =
Vtotal
= Vd0 cos α,
2
Vtotal = 2 × Vd0 cos α,
(1)
(2)
where, Vd0 is the ideal no-load voltage at the DC side;
α is the firing angle. If the rated capacity of a single
converter is SN , the reactive power of each six-pulse
converter is,
Q1 = Q2 =
Qtotal
= SN sin α,
2
Qtotal = 2 × SN sin α.
(3)
(4)
The power diagram of symmetrical control mode is
shown in Fig. 2(a) and (c), in which α1 = α2 = α.
Semicircle AB indicates the reactive power demand.
When Vtotal =0, the reactive power consumption reaches
the maximum, i.e. Qp.u. =1, because α1 = α2 = α=90◦ .
In sequential control mode, one of the six-pulse converters will be fully advanced or fully retarded, and
the other one provides the rest of the reference voltage.
Two cases are analyzed, as shown in Fig. 2(b) and (d).
Fig.2
Power diagrams of a two-converter in a series
Case 1: Vtotal > 0
V1 = Vd0 ,
(5)
V2 = Vtotal − V1 .
1148
YUAN Hongwen et al.: On the Sequential Control of ITER Poloidal Field Converters
Thus,
α1 = 0◦ ,
−V1
α2 = cos−1 ( Vtotal
).
Vd0
(6)
Case 2: Vtotal < 0
V1 = Vtotal − V2 ,
(7)
V2 = −Vd0 .
Thus,
−V2
α1 = cos−1 ( Vtotal
),
Vd0
(8)
Fig.3 Voltage allocation method of sequential control for
the two-converter in a series
α2 = 180◦ .
Once the firing angle of a six-pulse converter module
is fully advanced or fully retarded [9] , it means that this
converter will not consume reactive power, only consume or regenerate active power from or to the grid.
The power diagram of the reactive power consumption is shown in Fig. 2(b) and (d). Semicircles AO and
OB indicate the reactive power demand with sequential
control. Compared to that of symmetrical control, the
reactive power demand is greatly reduced, especially
when Vtotal =0, the reactive power consumption reaches
the minimum, i.e. Qp.u. = 0.
In practical cases, leakage inductance exists in the
converter transformer. If the impact of factors such
as commutation overlap angle and load current on the
reactive power is considered, the reactive power consumption of each six-pulse converter will be,
Fig.4 P &Q map of two six-pulse converters in a series
with sequential control (Id = IdN )
2.2
cos αi + cos(αi + γi )
Qi = Ud0
Id tan cos−1
2
Three converters in series
The topology of three six-pulse converters in a series is shown in Fig. 5. The total reference voltage is
allocated to the three converter modules. V1 , V2 , and
V3 represent reference voltages for converter modules
#1-#3, respectively. Ud1 , Ud2 , and Ud3 are the output
voltages of the converter modules.
cos αi + cos(αi + γi )
n
o2 ,
2
sin(2αi +2γi )−sin 2αi −2γi
[cos αi + cos(αi + γi )] +
2[cos αi −cos(αi +γi )]
r
(9)
where, i=1, 2, which represents the two six-pulse converters in a series; Qi is the reactive power consumption, αi and γi are respectively the firing angle and the
commutation overlap angle; Id is the load current. The
total reactive power consumption is the sum of Q1 and
Q2 ,
Qtotal = Q1 + Q2 .
(10)
The range of the firing angle is also limited by different factors, such as converter transformer impedance,
commutation overlap angle and load current [10] , so the
six-pulse converter modules are not allowed to be triggered with a firing angle of 0◦ or 180◦ . According to
the design requirement of the ITER PF converter system, each converter module is anticipated to output dc
voltage rated for ±1.05 kV. In this way, the voltage allocation principle of sequential control is illustrated in
Fig. 3. The reactive power diagram is shown in Fig. 4
when the load current is equal to the rated value.
Fig.5
Three converter modules in a series
For convenience of analysis, the impact of converter
transformer inductance is neglected; in a symmetrical
control case, each converter module shares 1/3 of the
1149
Plasma Science and Technology, Vol.16, No.12, Dec. 2014
total reference voltage.
Thus,
Vtotal
V1 = V 2 = V3 =
= Vd0 cos α,
3
(11)
Vtotal = 3 × Vd0 cos α.
(12)
α1 = α2 = α3 = α.
(13)
α1 = α2 = 0◦ ,
1 −V2
α3 = cos−1 ( VtotalV−V
).
d0
Case 2: −1/3 < Vtotal < 1/3
Thus,
The reactive power consumption of each six-pulse
converter is,
Q1 = Q2 = Q3 =
Qtotal
= SN sin α,
3
Qtotal = 3 × SN sin α.
V1 = Vd0 ,
V2 = Vtotal − V1 − V3 ,
V3 = −Vd0 .
(18)
α1 = 0◦ ,
1 −V3
α2 = cos−1 ( VtotalV−V
),
d0
◦
α3 = −180 .
(19)
Thus,
(14)
(15)
Case 3: Vtotal < −1/3
Fig. 6 illustrates the principle of the symmetrical and
sequential control method for three six-pulse converters
in a series upon the reactive power consumption. The
power diagram of three converter modules in symmetrical control mode is shown in Fig. 6(a)-(c). Semicircle AB indicates the reactive power demand. When
Vtotal =0, the reactive power consumption reaches the
maximum, i.e. Qp.u. = 1, because at this moment,
α1 = α2 = α=90◦ .
In sequential control mode, two of the three six-pulse
converters will be fully advanced or fully retarded so as
to reduce the reactive power consumption. As shown in
Fig. 6(b), (d), (f), semicircles AC, CB, BA indicate the
power diagrams of three series converters in sequential
control mode.
Case 1: Vtotal > 1/3
V1 = Vtotal − V2 − V3 ,
V2 = −Vd0 ,
V3 = −Vd0 .
(20)
2 −V3
α1 = cos−1 ( VtotalV−V
),
d0
◦
α2 = −180 ,
α3 = −180◦ .
(21)
Thus,
In practical cases, the impact of factors such as the
commutation overlap angle and the load current on the
reactive power consumption should be considered, so
the range of firing angle is limited so that the six-pulse
converter modules are not allowed to be triggered with
firing angle of 0◦ or 180◦ . With Eq. (9), the reactive
power consumption could be calculated. The total reactive power is the sum of Q1 , Q2 and Q3 .
V1 = V2 = Vd0 ,
(16)
Qtotal = Q1 + Q2 + Q3 .
V3 = Vtotal − V1 − V2 .
Fig.6
(17)
Power diagrams of three-converter in a series
1150
(22)
YUAN Hongwen et al.: On the Sequential Control of ITER Poloidal Field Converters
The voltage allocation principle of sequential control is shown in Fig. 7. Based on Eq. (22), the reactive
power diagram is illustrated in Fig. 8.
Fig.9
Schematic diagram of model for simulation
Fig.7 Voltage allocation method of sequential control for
three-converter in a series
Fig.8 P &Q map of three six-pulse converters in a series
with sequential control (Id =IdN )
3
Simulation results
The schematic diagram of the simulation model of
three series converter units is shown in Fig. 9. Simulations are performed based on SIMULINK to verify
the effectiveness of sequential control applied to multiseries converters for the purpose of reducing the reactive
power consumption of the converter system. In the simulation, the load current is replaced by a 27.5 kA DC
current source, which is the rated value of ITER PF
converters. The reactive power consumption is measured at the 66 kV busbar where the fundamental waveform is easier to be defined. The measured value mainly
represents the fundamental waveform reactive power.
When the simulation is performed, the reference voltage varies from −3.15 kV to 3.15 kV. In the case of two
series converters being simulated, the reference voltage
varies from −2.1 kV to 2.1 kV. The voltage allocation
method is described above. Meanwhile, symmetrical
control simulations are also made for a sharp contrast
with sequential control.
From the simulation results, as illustrated in Fig. 10,
it is obviously observed that, in symmetrical control
mode, the reactive power consumption reaches maximum when the active power consumption approaches
zero. On the contrary, reactive power consumption be-
(a) Two series converter modules with symmetric control,
(b) Two series converter modules with sequential control,
(c) Three series converter modules with symmetric control,
(d) Three series converter modules with sequential control.
Fig.10 Simulation results in the type of P &Q map
1151
Plasma Science and Technology, Vol.16, No.12, Dec. 2014
5
comes much lower in the case of sequential control.
For two six-pulse converters in a series, in symmetric control mode, the peak reactive power is as much
as 82 MVar; whereas in sequential control mode, the
value is 55 MVar, about 25 MVA reactive power being
saved. In the case of three six-pulse converters in a
series, the reactive power consumption is respectively
120 MVar and 95 MVar when active power approaches
zero. These results subsequently demonstrate the great
advantage of sequential control in improving the natural factor of multi-series converters, thus reducing reactive power consumption.
The views and opinions expressed herein do not necessarily reflect those of the ITER organization.
Acknowledgments
The authors also would like to express gratitude to
the staff of ASIPP for their helpful discussions.
References
1
In addition, it is necessary to note that the firing
angle is limited by many factors such as the leakage
inductance of transformer and stray parameters at the
ac busbar, which ultimately results in a commutation
overlap angle that means the firing angle cannot open
to 180◦ . In the ITER case, the range of the firing angle
of the PF converter under rated condition is (15◦ -135◦ ),
which should be taken into account in the evaluation of
the performance of sequential control.
4
Disclaimer
2
3
4
5
Conclusion
6
Compared to symmetrical control, sequential control effectively reduces the reactive power consumption
without any additional cost. The key point of sequential control is to allocate the reference voltage obtained
from the master controller to the series converter modules in a reasonable way.
7
The effect of sequential control is subjected to some
practical restrictions of normal operation, such as the
range of firing angle, which is limited by converter
transformer impedance, the commutation overlap angle and the load current. In that case, the firing angle
of thyrister the converter could not be fully advanced
or fully retarded. Based on the design requirement of
the ITER PF converter system and limitation of the
firing angle, a sequential control strategy is proposed.
Using this method, as much as 28.8% reactive power
consumption can be saved. The theoretical analysis
and simulation results are found in agreement, which
indicates the advantages of sequential control in reducing reactive power consumption.
8
9
10
Shimada M, Campbell D J, Mukhovatov V, et al. 2007,
Nuclear Fusion, 47: S1
Fu Peng, Gao Ge, Song Zhiquan, et al. 2008, Fusion
Science and Technology, 54: 1003
Fu Peng, Liu Zhengzhi, Xu Jia, et al. 2002, Fusion
Science and Technology, 42: 155
Benfatto I, Mondino P L, RoshZhi A, et al. 1995,
AC/DC Converters for the ITER poloidal field system. Proc. of the 16th SOFE, Champain, IL, USA
Xu L, Sheng Z, Fu P, et al. 2010, The Reactive Power
Compensation and Harmonic Filtering and the Overvoltage Analysis of the ITER Power Supply System.
Twenty-Fifth Annual IEEE Applied Power Electronics
Conference and Exposition-APEC , Palm Springs, CA
Mankani A D, Benfatto I, Tao J, et al. 2011, The ITER
Reactive Power Compensation and Harmonic Filtering
(RPC & HF) System: Stability & Performance. IEEE
24th Symposium on Fusion Engineering, Chicago, IL,
USA
Gaio E, Piovan R, Toigo V, et al. 1997, Bypass
Operation of the ITER AC/DC Converter for Reactive Power Reduction, Fusion Engineering. 17th
IEEE/NPSS Symposium, San Diego, California
Neumeyer C, Benfatto I, Hourtoule J, et al. 2013,
ITER Power Supply Innovations and Advances. IEEE
25th Symposium on Fusion Engineering (SOFE), San
Francisco, CA
Heo Hye-Seong, Park Ki-Won, Jeong In-Chol, et al.
Sequential Control of Small-scaled ITER Power Supply for Reactive Power Compensation. IEEE International Conference on Industrial Technology (ICIT),
Viña Delmar
Subhas Mukhopadhyay. 1978, IEEE Transaction on Industry Applications. IA-14: 594
(Manuscript received 17 February 2014)
(Manuscript accepted 25 March 2014)
E-mail address of YUAN Hongwen: yhw@ipp.ac.cn
1152