This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TASC.2020.2982877, IEEE
Transactions on Applied Superconductivity
THU-MO-PO4.12-01 1

Mobile Superconducting Magnetic Energy Storage

for On-Site Estimations of Electric Power System
Stability
Shinichi Nomura, Member, IEEE, Tanzo Nitta, and Takakazu Shintomi

Abstract—Renewable energy and electric power liberalization This work discusses a conceptual design study of the SMES
have become important watchwords for present electric power system for eigenvalue identification based on superconducting
systems. However, the stability of electric power systems is also engineering.
a serious concern. Superconducting magnetic energy storage
(SMES) systems are a promising candidate for the on-line The oscillation modes of a power system vary in real time,
evaluation of power system stability. From the viewpoint of the and their distribution is complex. In addition, the mobility
mobility of SMES systems, the authors carried out a design study of the SMES system is the most important parameter for
on a 1 MJ-class mobile SMES system using MgB2 Rutherford eigenvalue identification. In this work, the authors carried
cables. The results showed that 1) the 1 MJ-class mobile SMES
out a design study on a 1 MJ-class mobile SMES system
components can be installed in a 40 feet dry container; 2) because
of the effect of the force-balanced coil design, the SMES coil can whose components such as SMES coils, cooling systems, and
be excited up to 2.0 T or 3.0 T without reinforcements for the power converters are installed in a 40 feet dry container.
MgB2 Rutherford cables; and 3) the 1 MJ-class mobile SMES For the design flexibility of the stored energy of the SMES
coil can be cooled using 3 or 4 sets of conventional cryocoolers, system depending on the power system conditions, the authors
including the cooling system for a 80 K thermal shield at a
investigated the engineering feasibility of a 1 MJ-class SMES
cooling temperature of 20 K or 10 K. Therefore, the engineering
feasibility of the 1 MJ-class mobile SMES system using MgB2 coil using MgB2 Rutherford cables.
Rutherford cables is confirmed.
Index Terms—Eigenvalue, electric power liberalization, electric II. O N -S ITE E STIMATIONS OF E LECTRIC P OWER S YSTEM
power system stability, MgB2 , renewable energy, SMES. S TABILITY
A. Problems in the Present Electric Power System Stability
I. I NTRODUCTION
Fig. 1 shows a schematic diagram of the control methods for
ENEWABLE energy and electric power liberalization
R have become important watchwords for present elec-
tric power systems. However, the stability of electric power
power fluctuations in an electric power system. In conventional
electric power systems, these control methods can be realized
by controlling the output power or the excitation current of
systems is also a serious concern because of the lack of
the synchronous generators in hydro power plants or thermal
inertial energy of conventional synchronous generators and the
power plants. This work focuses on power fluctuations with a
limitations of the high-precision calculation of system stability.
cycle of less than 1 s, which is related to system stability. From
Since a low system stability may cause an unexpected power
the viewpoint of system stability, if the proportion of energy
oscillation with the natural frequency of an electric power
generated by renewable energy sources such as photovoltaic
system [1], [2], power system operation requires an excessive
power and wind power increases, the present electric power
margin to maintain a high system stability.
systems will have the following problems:
The objective of this work is to discuss the ability of a
superconducting magnetic energy storage (SMES) system to 1) Lack of inertial energy of conventional synchronous
identify the stability of electric power systems which is related generators, and
to the power fluctuation within a cycle of 1 s. The development 2) Limitations in the high-precision calculation of system
results of SMES systems include the commercialized system stability.
for bridging instantaneous voltage dips with a discharge time The first problem is related to a hardware problem in an
of less than 1 s [3]. From the viewpoint of electrical power electric power system. In conventional electric power systems,
engineering, Nitta et al. proposed the feasibility of using a the power fluctuations with a cycle of less than 1 s can be
SMES system for the direct identification of eigenvalues that absorbed by the inertial energy of the synchronous generators.
express the oscillation modes of a power system [4], [5]. However, because the photovoltaic power generation and wind
power generation are interconnected to the power system
S. Nomura is with Meiji University, 1-1-1 Higashimita, Tama-ku, Kawasaki,
Kanagawa 214-8571 Japan (phone: +81-44-934-7300; fax: +81-44-934-7909; through power converters, it may be difficult to maintain the
e-mail: [email protected]). stability of the electric power system due to the lack of inertial
T. Nitta was with the University of Tokyo (e-mail: [email protected]). energy of conventional synchronous generators. Therefore,
T. Shintomi is with High Energy Accelerator Research Organization, KEK,
1-1 Oho, Tsukuba, Ibaraki 305-0801 Japan (e-mail: [email protected]). the low system stability may cause an unexpected power
Manuscript received September 24, 2019. oscillation in the electric power system [1], [2].

1051-8223 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: University of Newcastle. Downloaded on April 14,2020 at 11:16:02 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TASC.2020.2982877, IEEE
Transactions on Applied Superconductivity
THU-MO-PO4.12-01 2

Large Economic load

Research target dispatching A B Power
G
Power fluctuation

control system
(System stability problems)
Load Generator
frequency
control Storage Small sinusoidal
system power oscillation
Governor-free Conventional target
Power system operation e.g. battery system
stabilizer (Power imbalance problems)
Small Self-regulation
(a)
1 sec. 1 min. 1 hr. Cycle

Fig. 1. Control methods for power fluctuations in electric power systems. Power Spectrum Storage system
For an effective interconnection of renewable energy sources, the main Natural
conventional research target is to improve the problem of power imbalance frequency
between power demand and power supply using battery energy storage
systems. On the other hand, the research target of this work is to enhance
the system stability with a cycle of less than 1 s.

time f0 fsm Frequency

On the other hand, the second problem is related to a
software problem in the electric power system. In general,
(b)
the system stability is evaluated by calculating the eigenvalue
that expresses the oscillation mode of the electric power
system. In conventional electric power systems, because the Half width
modeling of electric power equipment such as generators ∆P
and power transmission lines is relatively easy, the power
system can be operated according to the calculation of the
eigenvalue with a suitable margin. However, because electric
power liberalization is drastically accelerated in recent power Natural frequency f0 Frequency
systems, it is difficult to understand the properties of the power 䠄1䡚2 Hz䠅
equipment entirely. Therefore, to maintain a high system sta-
bility, the power system operation requires an excessive margin (c)
for the calculation of eigenvalue. Particularly, for renewable
energy systems, because the modeling of power equipment Fig. 2. System configuration (a) for the on-site identification of eigenvalue
using power electronics technologies has become an important in an electric power system. The storage system provides a small sinusoidal
research theme [6], a direct measurement technique is required power oscillation with a frequency, fsm . By measuring the electric power
disturbance at nodes A or B, the power oscillation depending on the natural
for the eigenvalue in an actual power system to confirm the frequency, f0 , of the electric power system is identified (b). The real part of
validity of the modeling results. the eigenvalue is obtained from the half width of the frequency response at
the natural frequency (c).

B. Identification Method for Electric Power System Stability SMES Current source
(Current source) converter Output current from the SMES
Fig. 2(a) shows a system configuration for the direct mea- Current (Unidirection) Direct supply to the grid
surement of the eigenvalue, λ, in the electric power system by controlling the converter.

[4], [5]. The storage system provides a small sinusoidal power Voltage
(Bidirection)
oscillation with a frequency, fsm , to the electric power system. Electric
Charge power
By measuring the electric power disturbances at nodes A and system
B, the power oscillation depending on the natural frequency, Discharge
Line
voltage
f0 , of the electric power system is identified, as shown in Fig.
2(b). The real part of the eigenvalue, Re (λ), is obtained from
the half width of the frequency response at the natural fre- Fig. 3. Current supply capability of SMES to the electric power system.
quency, as shown in Fig. 2(c). The real part of the eigenvalue is
related to the dumping effect against power fluctuation. When
current path of the power converter. Because of this feature,
the real part of the eigenvalue is small, an unexpected power
the primary advantages of the SMES system are as follows:
oscillation is easily caused by the low system stability.
To identify the eigenvalue, the storage system requires the 1) Current supply capability with a minimal influence on
controllability of charge/discharge with a rapid cycle of less the power system, and
than 1 s. Because the electric power system is a voltage source 2) Simple charge/discharge controllability with a rapid cy-
system, the grid-connected energy storage system should have cle.
the same capability of supplying current to the power system. Therefore, a SMES system is a promising candidate for
As shown in Fig. 3, the SMES system can directly provide eigenvalue measurement.
the output current to the power system by controlling the In a previous study, Yonezu and Nitta et al. investigated

1051-8223 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: University of Newcastle. Downloaded on April 14,2020 at 11:16:02 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TASC.2020.2982877, IEEE
Transactions on Applied Superconductivity
THU-MO-PO4.12-01 3

Discharge Oscillation mode 1

(Required energy: Esm) Generator mode 2
P(t)
Psm G G G G G

Pgrid
A B C D E

Charge Fig. 5. Example of power oscillation conditions.

0 t
III. C ONCEPTUAL D ESIGN S TUDY OF A M OBILE SMES
S YSTEM U SING M G B2 RUTHERFORD C ABLES
Fig. 4. Required energy of SMES for on-site identification of eigenvalue in
the electric power system. Pgrid and Psm are the operating power in the A. Design Considerations for the Mobility of SMES Coil
target power system and the peak power of the SMES system, respectively.
The authors carred out a design study on the SMES system
for the on-line estimation of power system stability. From
the viewpoint of the mobility of the SMES system, the
the feasibility of using the SMES system for the on-line
development target of this work is a 1 MJ-class SMES system
identification of eigenvalues using an analog simulator [5].
using MgB2 Rutherford cables. The primary advantages of
In this experiment, the SMES system is replaced by three AC
using MgB2 superconductors are as follows:
current amplifiers with a delta connection. The results obtained
by using 2 sets of generators with 40% of rated output power 1) Reduction of cooling power requirement,
and an infinite bus showed that the peak power of the SMES 2) System operation at 20 K,
system is almost 1% of the power capacity of the target power 3) Manufacturability of high-current conductors because of
system for eigenvalue identification. the effect of round strands, and
4) Design flexibility of the stored energy of the SMES
system depending on the power system conditions by
C. Required Specifications of the SMES System selecting an optimal cooling temperature.
In addition, from the viewpoint of mobility and the weight
Fig. 4 shows a schematic waveform obtained from the saving of the SMES coils, this study discusses the feasibility
SMES system for eigenvalue identification. The output power of using force-balanced coils (FBCs) [7], which are helically
of the SMES system, P (t), is wound toroidal coils. The features of FBCs are as follows:
P (t) = Psm sin ωsm t = Psm sin 2πfsm t, (1) 1) Balancing the electromagnetic forces by selecting an
optimal number of poloidal turn,
where Psm and fsm are the peak power and frequency of 2) Minimizing the mass of the structure for energy storage
the SMES system, respectively. Compared with the operating based on the virial theorem, and
power in the power system, Pgrid , the stored energy of the 3) Improvement of the stress limitation of superconductors
SMES system, Esm , for eigenvalue identification is given by in a high magnetic field range [8], especially in the case
of advanced superconductors such as MgB2 and high-
∫ Tsm
2 Psm temperature superconductors [9].
Esm = P (t)dt = , (2)
0 πfsm Here, the authors focus on the third feature of FBCs. From
the virial theorem, the required mass, M , of the structure for
where Tsm is the charge/discharge cycle of the SMES system. the electromagnetic forces is proportional to the stored energy,
According to (2), the SMES system requires 0.32 MJ of stored E, [10] as follows:
energy in the case of operation at 1 MW and 1 Hz, and 1.3 ρ
MJ of stored energy in the case of operation at 4 MW and 1 M = Qmax E, (3)
σ
Hz.
The mobility of a SMES system is the most important where Qmax is a shape factor, which is related to the maximum
parameter for eigenvalue identification. Fig. 5 shows an ex- tensile stress; and ρ and σ are the mass density and working
ample of the power oscillation modes between the generators. stress, respectively. On the other hand, the ampere-meters of
The oscillation modes of the power system vary in real time, the superconductors (current, I, × length, S), which is related
and their distribution is complex. From Fig. 5, the oscillation to the mass of the superconductor [11], are given by
modes at nodes A, B, and E can be easily identified. However, −1/3 −1/3
IS = µ0 Qis E 2/3 Bm , (4)
it may be difficult to identify the oscillation modes at nodes C
and D. At node C, the generator does not oscillate; however where Qis is a shape factor and Bm is the maximum magnetic
power oscillations occur among nodes B, D, and E. On the field. From (3) and (4), if the electromagnetic forces are
other hand, at node D, because oscillation mode 1 mainly supported by the tensile stress in the superconductors, the
occurs, mode 2 may be difficult to identify. Therefore, the working stress, σ, in the superconductors becomes
eigenvalue should be measured while changing the location of 1/3 Qmax
the SMES system. σ = µ0 Jop E 1/3 Bm
1/3
, (5)
Qis

1051-8223 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: University of Newcastle. Downloaded on April 14,2020 at 11:16:02 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TASC.2020.2982877, IEEE
Transactions on Applied Superconductivity
THU-MO-PO4.12-01 4

5.0
10-K Ic
4.0 property

0.42 m

Magnetic field (T)

20-K Ic
3.0 property
10-K operating
2.0 point
1.9 m 20-K operating point
1.0
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
Winding slot Load line of the SMES coil
000000000000000000000000000000000000000000000000000000000
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 3 turns x 28 layers 0.0
60 mm

00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
000000000000000000000000000000000000000000000000000000000
000000000000000000000000000000000000000000000000000000000 0 1000 2000 3000 4000 5000
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
MgB2 Rutherford cable
Current (A)
000000000000000000000000000000000000000000000000000000000
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 36 strands φ 0.83 mm
000000000000000000000000000000000000000000000000000000000
(18 x 2 strands)
60 mm Fig. 7. Operating points of the 1 MJ-class SMES coil and the critical current
2 mm

properties of the MgB2 Rutherford cable with 36 strands for the SMES coil.
The critical current properties were estimated from the catalog data provided
15 mm by Hyper Tech Research, Inc., and the test results of the MgB2 Rutherford
cable with 12 strands at 4.2 K [13].
Stabilizer (Copper sheet)
0.3 mm x 13 mm

data provided by Hyper Tech Research, Inc. (30 filaments with

Fig. 6. Schematic illustration of the 1 MJ-class SMES coil with the FBC
design for the on-line estimation of power system stability. The FBCs are
20% of the fill factor), the critical current reduced to 46% of
composed of 6 of helical coils with 6 poloidal turns. The darker hatch indicates the original critical current during the cabling process. Using
a complete helical winding. The MgB2 Rutherford cable is composed of 36 this reduction ratio, the critical current properties of the MgB2
strands with a strand diameter of 0.83 mm and an external copper sheet for
quench protection. The winding slot consists of 3 turns × 28 layers of the
Rutherford cable with 36 strands were estimated, as shown in
MgB2 Rutherford cables. Fig. 7.
From the design flexibility depending on the power system
conditions, the operating current variations of the SMES coil
where Jop is the operating current density of the supercon- were estimated as 1800 A at 2.0 T for a cooling temperature
ductor. The working stress, σ, is the mean value of the stress of 20 K, and 2700 A at 3.0 T for a cooling temperature of
in the structure [7]. 10 K. Although the load factors of the SMES coil increase to
This relationship implies that the SMES coil configuration approximately 80% of the critical current properties, as shown
with a low Qmax /Qis improves the stress limitation of the in Fig. 7, the load factor will decrease if the critical current
superconductors under the same energy, magnetic field, and properties of the MgB2 strand are improved in the future.
current density. Compared with toroidal field coils (TFCs) and The key parameters of the 1 MJ-class SMES coil are sum-
single solenoids, the FBCs can minimize the Qmax /Qis [8]. marized in Table I. In the 10-K operation case, the maximum
Therefore, FBCs can be expected to obtain a higher stored magnetic energy is 2.2 MJ with a coil current of 2700 A.
energy without reinforcements for the superconductors. This For quench protection, the magnetic energy is absorbed by
is one of the important properties of conduction-cooled coils the external dumping resistance. At a hot-spot temperature of
using cryocoolers. 150 K with a dumping voltage of 1500 V, the cross-sectional
area of the copper stabilizer should be more than 6.0 mm2
B. Design Parameters of 1 MJ-Class SMES Coil for quench protection. According to the catalog data provided
Fig. 6 shows the schematic illustration of the 1 MJ-class by Hyper Tech Research, Inc., the copper component is only
SMES coil with the FBC design. To reduce the total number 12% of the MgB2 strand. Therefore, the MgB2 Rutherford
of turns, the MgB2 Rutherford cable is composed of 36 strands cable includes an external copper sheet for quench protection,
with a diameter of 0.83 mm, wound onto the winding slot with as shown in Fig. 6. The dimensions of the external copper
3 turns × 28 layers. The minimum radius applied to the MgB2 sheet are 13 mm width and 0.3 mm thickness. The overall
Rutherford cable is 150 mm in the poloidal cross section. This current density of the MgB2 Rutherford cable including the
radius is allowable for the react and wind (R&W) method external copper sheet is 115 A/mm2 at a maximum magnetic
without critical current degradation on the design base [12]. field of 3.0 T.
Although a detailed analysis of the strain distribution in the The working stresses of the FBCs, TFCs, and single
MgB2 Rutherford cable is required, the FBCs will be wound solenoid with an aspect ratio of 4.0 were compared under the
by the R&W method. same magnetic energy, maximum magnetic field, and operating
Jimbo et al. developed MgB2 Rutherford cables and in- current density. In (5), the operating current density, Jop , is
vestigated the critical current degradation during the cabling calculated from the cross-sectional area of the MgB2 strands
process [13]. The MgB2 strand manufactured by Hyper Tech without the external copper sheet. The aspect ratios are defined
Research, Inc. (product no. 30-NM) has a diameter of 0.83 by the major radius/minor radius in the cases of FBCs and
mm and consists of 30 filaments. The critical current of the TFCs, and by the diameter/coil height in the case of single
Rutherford cable with 12 strands at 4.2 K was 1200 A under solenoid. As shown in Table I, the FBCs can reduce the
an applied magnetic field of 4.0 T. According to the catalog working stress in the MgB2 strands to 36% of that in the case

1051-8223 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: University of Newcastle. Downloaded on April 14,2020 at 11:16:02 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TASC.2020.2982877, IEEE
Transactions on Applied Superconductivity
THU-MO-PO4.12-01 5

TABLE I TABLE II
K EY PARAMETERS OF THE 1 MJ-C LASS SMES C OIL WITH THE FBC H EAT L OAD AND THE C OOLING S YSTEM OF THE 80 K T HERMAL S HIELD
D ESIGN U SING M G B2 RUTHERFORD C ABLES . FOR THE 1 MJ-C LASS SMES C OIL .

Major radius/Minor radius 0.72 m/0.18 m Heat conduction by a pair of the copper current leads 243 W
FBC helical windings 6 poloidal turns × 6 coils Radiative heat transfer 5.6 W
Total poloidal turns 6×6×84 turns Heat conduction from the coil supports 2.8 W
Winding method React & Wind Total heat load 251 W
Self inductance 0.60 H Number of cryocoolers∗1 2 sets
MgB2 Rutherford cable length 4.1 km *1 The heat capacity of the cryocooler is 175 W at 50 Hz.
Weight of the helical windings 7.6×102 kg
(Weight of the MgB2 strands∗1 ) (6.2×102 kg)
(Weight of the external copper sheet∗2 ) (1.4×102 kg)
Weight of the GFRP winding form∗3 4.2×102 kg 1) The outer and inner diameters of the coil supports were
Total weight of the SMES coil 1.2×103 kg 100 mm and 90 mm, respectively,
Cooling method Conduction cooling 2) The length of the coil support was 300 mm, and
Cooling temperature 20 K 10 K
Maximum magnetic energy 0.97 MJ 2.2 MJ 3) The SMES coil was supported by 6 sets of coil supports
Maximum magnetic field 2.0 T 3.0 T made of GFRP.
Coil current 1800 A 2700 A
Working stress in the MgB2 strands∗4 6.1 MPa 14 MPa
The load of the supports was less than 1/180th of the
(cf. TFCs case) (17 MPa) (38 MPa) withstand load when the allowable compressive stress was 240
(cf. single solenoid case) (28 MPa) (64 MPa) MPa.
*1 Mass density of the MgB2 strand is 7.7×103 kg/m3 . The AC losses in the SMES coil are mainly discussed in
*2 Mass density of the copper sheet is 9.0×103 kg/m3 .
terms of hysteresis loss in the MgB2 Rutherford cables. The
*3 Mass density of GFRP is 2.0×103 kg/m3 .
*4 The working stress is estimated from (5). conditions of the hysteresis loss estimation are:
1) The hysteresis loss is evaluated by the variation in the
center toroidal magnetic field of the SMES coil,
of TFCs and 21% of that in the case of single solenoid. The 2) The magnetization variation of MgB2 at 20 K is assumed
working stress in the FBC windings is significantly lower than as 0.0002-0.0006 T in the magnetic field range of 2.3-1.2
the allowable stress of the MgB2 strand without critical current T, and
degradation [14], even if the 99% Ic criterion for a cyclic stress 3) The critical current density at 10 K is assumed to be 2.5
[15] is applied to the MgB2 strand. Therefore, the 1 MJ-class times of that at 20 K.
SMES coil with the FBC design can be excited up to 2.0 T or
Table II shows the heat loads of the 80 K thermal shield
3.0 T without reinforcements for the MgB2 Rutherford cable.
for the 1 MJ-class SMES coil. Although the total heat load
This is an important property for conduction-cooled coils using
is mainly caused by heat conduction from the 2700 A copper
cryocoolers.
current leads, the 80 K thermal shield can be cooled using 2
sets of conventional cryocoolers.
C. Heat Load Estimations and Cooling System Table III summarizes the heat loads of the 1 MJ-class
This section discusses the heat loads of the 1 MJ-class SMES coils. Cases 1 and 2 show the heat loads of the 20-
SMES coil and evaluates the applicability of the conduction K operation cases. The peak power of the SMES system
cooling method using conventional cryocoolers. is 1 MW at 20-K operation. On the other hand, cases 3-8
The design conditions of the current leads are as follows: show the heat loads of the 10-K operation cases. When the
1) A current capacity of 2700 A, cooling temperature was reduced to 10 K, the peak power of
2) Copper current leads were used between 300 K and the the SMES could be increased up to 4 MW. Because of the
80 K thermal shield. The heat conduction by the copper effect of the HTS current leads, the heat conduction by the
current lead was 45 mW/A [16], and current leads can be reduced. The total heat load is mainly
3) High-temperature superconducting (HTS) current leads caused by AC loss (hysteresis loss) in the MgB2 strands
were used between the 80 K thermal shield and the during the charge/discharge operations. The results showed
SMES coil. The heat conduction by the HTS current that the SMES coil can be cooled using 1 or 2 sets of
lead was 120 mW/kA [17]. conventional cryocoolers even when the cooling temperature
The conditions of the radiative heat transfer calculation are is 10 K. Therefore, including the cooling system for the 80 K
as follows: thermal shield, the 1 MJ-class SMES coil requires 3 or 4 sets
of cryocoolers.
1) The surface area of the SMES coil torus was 60 m2 ,
2) Using the emissivity of aluminum, the effective emis-
sivities were calculated as 0.020 in the range of 300 K D. Schematic Layout of Mobile SMES System
to 80 K, and 0.016 in the range of 80 K to the cooling For the on-site estimations of power system stability, the 1
temperature (20 K or 10 K), and MJ-class mobile SMES system can be installed in a 40 feet
3) The superinsulation effect is 1/10th of the heat load dry container, as shown in Fig. 8. As shown in Tables II and
without superinsulations. III, in most cases, the 1 MJ-class SMES system requires 4
The heat conduction from the coil supports was estimated sets of conventional cryocoolers, including the cooling system
under the following conditions: for the 80 K thermal shield. Because the dimensions of the

1051-8223 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: University of Newcastle. Downloaded on April 14,2020 at 11:16:02 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TASC.2020.2982877, IEEE
Transactions on Applied Superconductivity
THU-MO-PO4.12-01 6

TABLE III
H EAT L OADS AND THE C OOLING S YSTEM OF THE 1 MJ-C LASS SMES C OIL WITH THE FBC D ESIGN U SING M G B2 RUTHERFORD C ABLES .

Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 case 7 case 8

Output power of SMES Psm (MW) 1.0 1.0 2.0 2.0 3.0 3.0 4.0 4.0
Frequency (charge/discharge cycle) fsm (Hz) 1.0 2.0 1.0 2.0 1.0 2.0 1.0 2.0
Available energy Esm (MJ) 0.32 0.16 0.64 0.32 0.95 0.48 1.3 0.64
Charge/discharge ratio∗1 (%) 33 16 29 15 44 22 58 29
Maximum/minimum magnetic field∗2 (T) 1.5/1.2 1.5/1.4 2.3/1.9 2.3/2.1 2.3/1.7 2.3/2.0 2.3/1.5 2.3/1.9
Cooling temperature (K) 20 20 10 10 10 10 10 10
Heat conduction by a pair of the HTS current leads (W) 0.65 0.65 0.65 0.65 0.65 0.65 0.65 0.65
Radiative heat transfer (W) 0.23 0.23 0.23 0.23 0.23 0.23 0.23 0.23
Heat conduction from the coil supports (W) 0.25 0.25 0.28 0.28 0.28 0.28 0.28 0.28
AC loss (Hysteresis loss) (W) 1.7 1.3 3.4 2.2 5.4 3.4 7.7 6.9
Total heat load (W) 2.9 2.4 4.6 3.3 6.6 4.5 8.8 8.0
Number of cryocoolers∗3 1 set 1 set 1 set 1 set 2 sets 1 set 2 sets 2 sets
*1 The charge/discharge ratio is defined as the available energy/the maximum magnetic energy (See Table I.)×100.
*2 The center toroidal magnetic field variation during the SMES operation.
*3 The heat capacities of the cryocooler at 50 Hz are 40 W in the 20-K case and 5.4 W in the 10-K case, respectively.

conventional compressor for the cryocooler are almost 0.5 40 feet dry container
2m 2m
m × 0.5 m, the cooling system for the SMES coil will be Height: 2.386 m

installed in an area of 2 m × 2 m including chillers for the

1.2 m
2.352 m
Cooling Power converter

2m
2m
compressors. Fig. 8 shows the dimensions of the conventional system
power converter for photovoltaic (PV) power systems as an 5m

example. The detailed configuration of the current source 12.033 m

power converter for the SMES coil will be discussed in a MJ class SMES coils Refer to a 3.2 MVA conventional
future work.
Force-balanced coils power conditioner for PV systems
Because the required power capacity of the SMES system
MgB2 Rutherford cable

Cryocooler (10 K or 20 K) Compressors and chillers for the cryocoolers
is almost 1% of the power capacity of the target electric power
system [5], the feasible examples of the installation sites of
Fig. 8. Schematic layout of the 1 MJ-class mobile SMES system for on-line
the 1 MJ-class mobile SMES system shown in Fig. 8 are as estimations of power system stability. From the viewpoint of mobility, the
follows: SMES coils, cooling system, and power converter are installed in a 40 feet
dry container.
1) In the case of 100 MVA-class substations, the 1 MJ-class
SMES system can be used as an eigenvalue-measuring
device. The eigenvalue measurement results indicate
that, if the power system requires an enhancement in 3) The 1 MJ-class mobile SMES coil can be cooled using
system stability, the SMES system can be used as a 3 or 4 sets of conventional cryocoolers, including the
power system stabilizer by increasing the number of 1 cooling system for the 80 K thermal shield even when
MJ-class SMES units, and the cooling temperature is 10 K.
2) In the case of 1 MW-class solar power plants, the 1
MJ-class SMES system can be used as an eigenvalue- Therefore, the engineering feasibility of the 1 MJ-class mobile
measuring device by reducing the output power of the SMES system using MgB2 Rutherford cables was confirmed.
SMES system to 1% of the rated power. In addition, it The future works are as follows:
can be used as a power system stabilizer by the rated
power operation. 1) Investigation of the effective installation sites for eigen-
value identification through power system analysis,
IV. C ONCLUSIONS 2) Evaluation of the re-cooling time interval and optimiza-
tion of the cooling scenario to achieve full mobility of
A SMES system is a promising candidate for the on-line the SMES system using cryocoolers, and
evaluation of power system stability. From the viewpoint of 3) Experimental verification of the design study results by
the mobility of a SMES system, the authors carried out a developing a laboratory prototype using MgB2 Ruther-
design study on a 1 MJ-class mobile SMES system using ford cables.
MgB2 Rutherford cables. The following conclusions are drawn
from the results: The system stability can be evaluated by measuring the
1) For the on-site identification of eigenvalue, the 1 MJ- eigenvalue. However, the required energy of SMES cannot be
class mobile SMES components can be installed in a 40 concretely identified to maintain a high system stability. The
feet dry container, required energy of the SMES system can be estimated from
2) Because of the effect of the FBC design, the SMES the eigenvector of the power system. The final objective of
coil can be excited up to 2.0 T or 3.0 T without this work is to establish an on-line identification technique for
reinforcements for the MgB2 Rutherford cables, and the eigenvector using a SMES system.

1051-8223 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: University of Newcastle. Downloaded on April 14,2020 at 11:16:02 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TASC.2020.2982877, IEEE
Transactions on Applied Superconductivity
THU-MO-PO4.12-01 7

ACKNOWLEDGMENT
The authors would like to thank RASMES (Research Asso-
ciation of Superconducting Magnetic Energy Storage) in Japan
for their valuable discussions and collaborative work.

R EFERENCES
[1] K. Prasertwong, N. Mithulananthan and D. Thakur, “Understanding
low frequency oscillation in power systems,” International Journal of
Electrical Engineering Education, vol. 47, no. 3, pp. 248–262, Jul. 2010.
[2] J. Ma, P. Zhang, H. Fu, B. Bo, Z. Dong, “Application of phasor measure-
ment unit on locating disturbance source for low-frequency oscillation,”
IEEE Trans. Smart Grid, vol. 1, no. 3, pp. 340–346, Dec. 2010 (Refer to
ResearchGate).
[3] S. Nagaya et al., “The state of the art of the development of SMES
for bridging instantaneous voltage dips in Japan,” Cryogenics, vol. 52,
pp. 708–712, 2012.
[4] T. Nitta, Y. Shirai, Y. Ito, “Evaluation of steady state stability of electric
power system by use of superconducting magnet energy storage,” IEEJ
Trans. Power and Energy, vol. 116-B, no. 6, pp. 678–684, 1996 (in
Japanese).
[5] T. Yonezu, T. Nitta and J. Baba, “On-line identification of real parts
of eigenvalues of power system by use of superconducting magnetic
energy storage,” IEEE Power and Energy Society General Meeting -
Conversion and Delivery of Electrical Energy in the 21st Century, DOI:
10.1109/PES.2008.4596943, 2008.
[6] L. Hui et al., “Damping control strategies of inter-area low-frequency
oscillation for DFIG-based wind farms integrated into a power system,”
Electrical Power and Energy Systems, vol. 61, pp. 279–287, 2014.
[7] S. Nomura and H. Tsutsui, “Structural limitations of energy storage
systems based on the virial theorem,” IEEE Trans. Appl. Supercond.,
vol. 27, no. 4, Art no. 5700106, Jun. 2017.
[8] S. Nomura, Y. Ohata, T. Hagita, H. Tsutsui, S. Tsuji-Iio, and R. Shimada,
“Helically wound coils for high field magnets,” IEEE Trans. Appl.
Supercond., vol. 14, no. 2, pp. 709–712, Jun. 2004.
[9] S. Nomura, S. Akita, R. Shimada and T. Shintomi, “Design considerations
for SMES systems using MgB2 and/or high-temperature superconduc-
tors,” IEEE Trans. Appl. Supercond., vol. 16, no. 2, pp. 590–593, Jun.
2006.
[10] Y. M. Eyssa, R. W. Boom, “Considerations of a large force balanced
magnetic energy storage system,” IEEE Trans. Magn., vol. 17, no. 1,
pp. 460–462, Jan. 1981.
[11] R. W. Moses, Jr., “Configurational design of superconductive energy
storage magnets,” Advances in Cryogenic Engineering, 21, C-2, pp. 140–
148, 1975.
[12] T. Yagai et al., “Development of design for large scale conductors and
coils using MgB2 for superconducting magnetic energy storage device,”
Cryogenics, vol. 96 pp. 75–82, 2018.
[13] M. Jimbo et al., “Investigation on Ic degradation of MgB2 Rutherford
cables by deformation during cabling process,” IEEE Trans. Appl. Super-
cond., vol. 29, no. 5, Art no. 8003305, Aug. 2019.
[14] H.-S¿ Shin, M. A. Diaz, Z. Bautista, H. Oguro and S. Awaji, “Charac-
terization of electromechanical properties in differently sheathed MgB2
wires under uniaxial tension,” IEEE Trans. Appl. Supercond., vol. 29,
no. 5, Art no. 8400805, Aug. 2019.
[15] K. Osamura, S. Machiya and G. Nishijima, “Reversible stress and strain
limits of the critical current of practical REBCO and BSCCO wires,”
Supercond. Sci. Technol., vol. 29, Art no. 094003, 2016.
[16] Y. Iwasa, Case studies in superconducting magnets, Design and op-
erational issues, Second edition. New York: Springer Science+Business
Media, LCC, 2010, p. 274.
[17] Y. Yamada et al., “HTS current leads prepared by the TFA-MOD
processed YBCO tapes,” IEEE Trans. Appl. Supercond., vol. 21, no. 3,
pp. 1054–1057, Jun. 2011.

1051-8223 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: University of Newcastle. Downloaded on April 14,2020 at 11:16:02 UTC from IEEE Xplore. Restrictions apply.