| By Rajat Khanna, Vineet
Jain ABSTRACT
A design
feasibility study has been conducted for a 240
MVA, high temperature superconducting grid auto-transformer.
Conclusions are relevant to superconducting power
transformers in general. Economic benefits may be
achieved subject to assumptions about achievable
conductor properties, cost of components and
power system operating requirements. Liquid
nitrogen cryogenics is relatively cheap and
simple and refrigeration power demand is reduced
by a factor of the order of 20 compared to the
low temperature case (LTS). Attention is drawn to
the importance of AC losses in the superconductor
and the difficulty of keeping these sufficiently
low. Various technical problem areas and their
likely influence on the overall design concept
are reviewed. Three particularly important
influences are identified: insulating properties
of nitrogen coolant, required transformer
performance in the through-fault conditions and
mechanical strength to withstand electromagnetic
forces. Typical conductor design and winding
configuration result in an efficiency of the
superconducting transformer of 99.85 % .The
savings in the cost of losses over the life of
transformer provide the economic-feasibility of
this design approach.
INTRODUCTION:
From the moment
of discovery of the high magnetic field , high
current density superconductors, engineers have
been challenged to incorporate this " loss-less"
conductor in power equipment . These attempts
have met with some success in the case of DC
applications such as in DC field windings of
generators, because of the loss behavior of these
materials in time varying magnetic fields ,AC
applications have been hindered . One such
application is in power transformers . Work from
1961 to 1977 generally indicated that
superconducting transformers were not feasible .
It was earlier
examined what was necessary to make a
superconducting transformer viable and concluded
that to minimize flux jumps, the magnetic field
intensity at the windings must be kept low ;
therefore inter leaving the high voltage windings
was proposed and the iron core should be operated
large volume for the device . After that 570 MVA
transformer made with superconducting tapes and
interleaved windings was studied .This effort was
followed by examining a number of configurations
including the elimination of iron. These
approaches proved to higher loss. Later it was
rechecked the question of whether interleaving
was necessary. Then 600 MVA superconducting
transformer made up of Nb3Sn tapes was
examined. A unique concept, later examined called
for a set of superconducting windings enclosed by
a set of 77 K auxiliary copper windings. This
concept also proved uneconomical because of large
volume .In 1977 superconducting transformers
using conductors operating in the low AC loss
"Meissner" region .This design proved
uneconomical because of the overall transformer
volume .
These failures
to devise viable techniques for applying
superconductors to transformers were due to large
volume configurations, and high AC losses due to
large AC magnetic fields or large superconductor
volumes.
In 1981, a 1000
MVA ,500/22 KV line-to-line ,1300 BIL low
temperature superconducting power transformer (LTS)
which uses superconducting windings . The design
uses a multiple winding configuration which is
current limiting under fault conditions .This
conductor design and winding configuration
resulted in an efficiency of superconducting
transformer of 99.85 % . The savings in the cost
of losses over the life of the transformer
provide the economic feasibility of this design
approach.
The advent of
high temperature superconducting (HTS) materials
has renewed interest in the possibilities for
superconducting power apparatus offering real
economic benefits, within power ratings typical
of present system practice. Previously developed
low temperature superconductors (LTS) require
cooling by liquid helium about 4.2K, with
advanced cryogenic technology that is expensive
both in terms of cost and refrigeration power
expended per unit of heat power removed from the
cryostat. The technology for the new materials,
which is based on liquid nitrogen (LN2)
at temperatures up-to about 79K, is simpler and
cheaper. The ratio of refrigeration power used to
heat removed is also reduced from around 500 to
about 25.
PROPERTIES
OF HTS MATERIALS:
The most
promising known HTS materials for high current
applications are given below:
Yttrium
compounds ( YBCO)
Y1Ba2Cu3O7-x
Tc=92
Y2Ba4Cu7O15-y
Tc=95K
Bismuth
compounds (BISCCO)
Bi2Sr2Ca1Cu2Oy
Tc=80K
Bi2-xPbxSr2Ca2Cu3Oy
Tc=110K
Thallium
compounds
(TlPb)1Sr2Ca2Cu3O9
Tc=120K
Tl2Ba2Ca2Cu3O10
Tc=125K
Mercury
compounds
Hg1Ba2Ca2Cu3O10
Tc=153K
Since the
discovery of HTS , the development of the
processing technology has resulted in a steady
increase in the critical current density in short
and long lengths of tapes and wires based on
BISCCO and YBCO superconductors . Critical
currents in the region of 400-to-800A/mm2
and upto 200A/mm2 at 77K and zero
fields have been obtained in short KM lengths of
Bismuth tapes, respectively. For use at lower
temperatures and very high fields, Bismuth tapes
have shown to be able to carry in excess of 20000A/mm2
at 4.2K in a field of 20T.
Recent
developments at Los Alamos and Oak Ridge on the
processing of YBCO tapes (IBAD and RABIT
techniques) have demonstrated the potential of
this superconductor achieving critical currents
of the order of 1000A/mm2 in short
lengths of tapes. Twisted multifilamentry tapes
are also being developed by several manufactures
to reduce the hysteretic losses in AC power
applications.
Specification
for required HTS :
The critical
current density is 80A/mm2 RMS (referred
to the total conductor cross section), achievable
at a maximum temperature of 79K in a magnetic
field of 0.8T (parallel to the tape surface),
with acceptable AC conductor losses. These form
the basis of the characteristics that are
specified in the design. It is to be noted that
the working RMS current density in a conventional
transformer winding is around 3A/mm2
and about 12A/mm2 in direct water
cooled winding.
However, the
important advantage of superconductivity for a
power transformer lies not so much with the
increase in working current density as with the
saving of ohmic losses.
AC loss
characteristics:
One of the most
important issues concerning power application of
superconductors is the AC loss at 50Hz and 60Hz.
Although the resistive loss is virtually zero in
superconductors, hysteretic loss arises in an
alternating magnetic field. In real application,
the losses are produced by both the self-field as
in a cable and the external field as in a coil.
Compared with LTS, the operation of HTS at LN2
temperatures means a great reduction in the
refrigeration cost and a significant improvement
in the cryogenic stability for the device.
Nevertheless, the problem of losses in the
superconducting devices remains the dominating
factor for the viability and the success of the
HTS in power engineering application.
The calculation
of AC losses for the design assumes two
components: the hysteretic loss in 5µm thick
superconducting sheets, and the coupling losses
due to currents in the silver matrix. For this
design, using the above conductor specification,
the hysteretic and coupling current losses are of
the same order. Above Fig. shows the AC loss in
the superconducting core of a particular tape as
a function of the peak AC field, with the field
dependence of the critical current taken into
consideration. It can be seen that the loss
increases with B03 at
fields below 10mT, where the tape is fully
penetrated. Above this field, the loss increases
linearly with B0.
Design
aspects of HTS transformer:
Overall
concept and principle parameters:
The principal
features of the design are the removable of the
copper windings and their replacement by HTS
equivalents. These are only a fraction (less than
10%) of the bulk of conventional windings. The
inevitable result is windings of reduced
mechanical strength, which will stand neither the
radial bursting forces nor the axial compressive
forces that occur during through- fault
conditions without special strengthening
structural features. The tap winding is kept
outside the cryostat to avoid the heat which
could otherwise flow into liquid nitrogen through
several connections.
With bulk of
ohmic losses (in common and series windings )
removed , it is possible to cope with the core
loss and remaining ohmic loss(in the tap windings)
by forced gas cooling, leading to an oil less
design of transformer. This has a great advantage
of reducing fire risk and environmental hazard
from oil spillage. The superconducting winding
have small thermal mass and are cryogenically
stable (i.e. capable of returning to normal
operation after a period of abnormal heating
without disconnecting and cooling down) over only
a small range of temperature rise. In
consequence, the HTS design has very little
capability to recover from a through-fault
condition without disconnection, in contrast to a
conventional transformer. This is the chief
weakness of the HTS design.
Note, however,
that the HTS design can survive a through fault,
although it subsequently needs disconnecting for
a period, and there is also good overload
capability.
To ensure
efficient cooling, it is necessary to have direct
contact between the LN2 coolant and
the conductor. However, this contact is only
required over a small percentage of the total
surface area and take place on a part of the
edges of the conductor strip forms the winding
discs. This requirement has implications for the
manner in which the inter-turn insulation is
arranged which must leave the edges bare.
Balance of
electric and magnetic loading (i.e. total ampere
turns v/s core flux) is one fundamental freedom
in the design process. With a superconducting
transformer, a natural tendency would be to
increase electric loading as far as possible,
since it is lossless and economic in space
requirement, and correspondingly reduce magnetic
loading. Thus, the tendency would be to minimize
radial clearances between windings, consistent
with the necessary voltage withstand capability ,
in order to minimize the permeance of leakage
flux paths, and so permit the greatest electric
loading consistent with the reactance level.
However, in
basic HTS design, another consideration takes
precedence. This is the capability to withstand
mechanical forces, which are proportional to the
square of electric loading and also tend to
increase as size is reduced. The overall
consequence is that the total ampere turns have
been kept the same as in a typical conventional
design. The radial space between common and
series windings has been deliberately increased
beyond the minimum demanded by voltage
considerations, in order to adjust the leakage
reactance.
The dielectric
strength of LN2 is very good, but that
of gaseous nitrogen is less. In HTS design, the
nitrogen in the cryostat is in the near saturated
state and boiling may occur with the release of
gas bubbles into the liquid.
The design
features of the HTS transformer are described
below.
Core:
This is similar
to a conventional design, but with leg length and
window spans reduced and consequent saving in
core weight and overall size.
Tap
change arrangements:
The copper tap
windings are retained, adjacent to the core leg,
outside the cryostat connected to a tap-changer,
both of conventional design.
Cooling
of core and tap windings:
Forced-convection
gas cooling in which nitrogen is used. Gas is
forced through axial ducts (parallel with legs)
to cool the core legs and tap windings. Top and
bottom yokes are fitted with fiberglass cowling
to contain fanned gas and direct it over yoke.
Common
and series windings:
Windings are
constructed of rectangular section composite
conductor, probably comprising about 33%
superconducting fibers and 67% matrix metal (silver
or silver alloy).The inter-turn insulating tape
is applied to one face of the conductor.
Individual tapes making up the conductor will be
transposed in a normal way. Discs at the high
voltage ends of the series windings are
interleaved in pairs.
Winding
reinforcement:
The outer
diameter of the series (high voltage) winding is
reinforced with fiberglass hoops (possibly pre-stressed)
or a continuous cylinder. A substantial inner
cylinder with multiple spacing sticks supports
the inner diameter of the common winding.There
are annular clamping plates at the top and bottom
of the complete in-cryostat winding assembly,
pulled together through bolts or studs, all
constructed of insulating material.
Cryostat:
One cryostat
per leg is used, with cryostat comprising a
vacuum vessel constructed double-skinned
fiberglass, with the vacuum continuously pumped.
Cryostat pressure will be slightly in excess of 1
bar, containing nitrogen at 77K. The intermediate-voltage
leads pass through the top lead. The high voltage
lead passes centrally through the cryostat wall.
Fault and
overload capability:
For any
substantial through fault, disconnection is
required and a period of minutes is needed before
reconnecting. The transformer can survive the
most severe through-fault for about 170 ms,
within which time disconnection must be achieved.
Internal faults are sensed by in-built monitors
or terminal voltage and current sensors, which
initiate disconnection. Over load capability is
expected.
Housing:
Oil-less design
obviates the need for a tank, which is replaced
by a steel structure carrying the load exerted by
bushings. External housing is required for
weather-proofing and a gas-seal for the forced-convection
nitrogen coolant and acoustic noise reduction.
The principal
parameters of the transformer design are listed
below.
kVA: 240 000
Normal volts:
400/132kV
Tappings: 132kV+15%-5%
in steps (+/-neutral end)
Line currents
at normal volts: 346/1054A
Diagram no: Yy0
Auto
Guaranteed
reactance: 20%
Rated current
densities: series windings=39.6A/mm2,
common
Winding=36.9A/mm2(39.6A/mm2,
minimum tap setting)
Tap winding=3.0
A/mm2(conventional)
Between the
tapes there are thin insulation strips. The edges
are partially exposed to the coolant. The area of
the Superconducting exposed to the coolant is 17%
of the total edge area for the common winding and
33% for the series winding (limits for both
values are due to the required mechanical
strength) both windings are supported by
uniformly spaced axial sticks covering
approximately 25% of the inner surface area.
Thermodynamic
stability:
HTS material is
inherently much more stable against local
temporary effects than LTS material; higher
specific heat makes it far less sensitive to
transient temperature excursion; and the size of
the region that is adiabatically stable is much
greater. If, due to malfunction, quenching occurs
during normal operation, there is ample time (over
15s) to disconnect the transformer before
irreversible damage is caused. However, thermal
capacity of the windings is much lower than in a
conventional transformer.
Dielectric
breakdown:
By using LN2
as a coolant it automatically becomes a part of
insulation system. Further, because this coolant
can be in intimate contact with the conductor, it
becomes the principal insulant for the main
windings. The choice of gas cooling for the core,
tap winding and connection leads means that the
conductor insulation in the area must rely on
solid material. It is proposed that the gas is
dry nitrogen.
As a gas,
nitrogen obeys Paschens law and increases
in break-down strength as the temperature is
decreased towards its liquid point. For gaps of
10mm and over, gaseous nitrogen has breakdown
strength of 6.4kV/mm RMS at 50 Hz AC, with
temperature of 77K and pressure of 1 bar. The
voltage withstands properties LN2
mixed with nitrogen gas bubbles may be assumed to
be determined by the properties of the gas. It is
concluded that all insulating parts of the
transformer solid , liquid or gas have a
breakdown strength equal to or greater than 6.4kV/mm
RMS AC .
Through-fault
condition:
Through-fault
condition poses a serious problem. The basic
design will recover ( without disconnection )
from a through fault of 2 p.u. current plus
maximum offset transient , provided that the
fault is cleared within 64ms ; this is just on
the edge of possibility for fast acting
protection in favorable circumstances.
Table:
Recovery time
and survival time, as a function of fault level
| Fault level,all
with doubling offset transient |
2 p.u. |
3 p.u |
4 p.u. |
5 p.u. |
| Maximum duration
allowing recovery , ms |
64 |
28 |
16 |
10 |
| Survival time
,ms |
1040 |
460 |
260 |
166 |
Higher fault
level call for disconnection of the transformer,
and will then need to remain disconnected for
substantial fraction of one hour in order to cool
down ready for resumption of normal working. This
disconnection must be achieved in a certain
maximum time, if the transformer is to survive
the fault and be fit for reconnection in due
course. With highest possible fault level (5 p.u.current)
, the survival time in the fault condition is
about 170ms . These times could be improved using
sub-cooled LN2.
Electromagnetic
Forces:
It is well
known that the forces acting on transformer
windings in the through-fault condition are very
high and can cause structural failure. There are
three principal forces to consider: an axial
compressive force Sc; an axial
displacement force Sd which, if there
is an initial axial displacement between the
center of symmetry of two windings, acts to
increase that displacement; and a radial force Sw,
which is a bursting force on the outer winding (that
must be withstood by tensile hoop force Sh ,
where Sw = 2Sh), and is a
buckling force on the inner winding .
Sc
tends to produce axial compressive failure of the
winding structure, with the individual conductor
sections of the winding disc yielding and
zigzagging out of their axial/radial alignment. Sd
can overpower the clamping arrangements that hold
the primary and secondary windings in axial
relation to each other, and in extreme cases can
lead to progressive collapse. It is modern to
construct the transformer with sufficient
attention to symmetry to ensure that the initial
displacement is minimized so that Sd
is not a problem .Sw can cause the
outer winding structure to yield in hoop tension
and the inner winding to deform into a clover -leaf
pattern with the winding collapsing radially
inwards between adjacent axial spacing sticks.
Therefore in
this design with increased current density in the
superconductor ,the axial forces Sc
and Sd have to be sustained on a much
smaller area of the windings.
Cryostat and
cooling:
To provide the
required working temperature, three independent
toroidal cryostats are designed to contain the
three phases. The cryostat must be non metallic
because of eddy currents that will otherwise be
induced in it by surrounding magnetic fields.
Total heat leaking from the room temperature
environment should be kept to a fraction of
winding losses. Owing to the limitation imposed
on thermal insulation thickness by window
dimensions, solid insulation is ruled out by the
excessive leak losses through the wall. Another
source of heat leak is the conductor entry ports
(or thermal bushings), which connects the low
temperature superconducting windings with the
room temperature leads. One HV port and three LV
ports per cryostat are needed. For the high
voltage connection, cable technology will be used
to design a link through the cryostat wall to the
bushing at a point half-way up the cryostat, As
shown in fig.(this is also convenient because the
series winding is formed of two parallel circuits
which can meet at the centre of the winding
height).This will leave the top clear for the
other connections to exit the cryostat and for
the tap winding connections to be brought out of
the tap changer with the large clearances
necessary to satisfactorily insulate them in dry
nitrogen gas.
Protection
requirements:
The normal main
protection would be required substantially
without change, but there are three important
aspects that require special consideration.
First, the HTS transformer does not have
conventional capability to withstand any through-fault
for upto 3s. Means must therefore be provided to
initiate disconnection when through-faults exceed
a permissible level of severity, which is
a combined function of current magnitude and time.
A suitable form of function can be built into a
simple intelligent controller, which is
continuously sampling the currents in all three
phases and which triggers disconnection if the
maximum permissible integrated value of the
function is exceeded.
Secondly, the
possibility will always exist of a local part of
the winding quenching due to unforeseen
circumstances, such as local structural looseness
leading to vibration and over heating. An
acoustic sensor is also employed to detect noise
in the LN2.
Finally, steps
must be taken to reduce in-rush current to
acceptable levels, e.g. by synchronized switching
or by using external resistors or fault current
limiters.
Loss
analysis:
| |
HTS |
Conventional |
| Core loss |
8 |
9 |
| Clamp stray loss |
5 |
5 |
| Tank loss |
- |
7 |
| Total copper
loss |
<1(tap) |
79 |
| Refrigerator
power |
7 |
- |
| Gas-cooling fan
loss |
2 |
- |
| Estimated total
loss |
23 |
100 |
Total loss of
conventional design=100%
Comparative
assessment of HTS and conventional transformer:
Above table
summarizes the total losses of the HTS
transformer design and compares them with the
conventional reference design. Losses
are expressed in percentage form, with the total
loss in conventional design taken as 100%. Table
below shows the significant global features,
covering size and construction as well as
performance. On the construction side, the main
points of interest are the reduction in core size
and weight for the superconducting transformer,
the elimination of a large weight of copper, and
oil less cooling. It is interesting to observe
that the reduction of winding size enables the
overall size of the transformer to be reduced,
despite needing room for the cryostats around
each core leg. Thus the design concept has the
effect of easing transport problems and makes a
significant contribution to the reduction of
environmental risks and fire hazards.
Table:
Comparison of technical features
| Parameter |
HTS |
Conventional |
| Core length |
88.5 |
100 |
| Height |
82.4 |
100 |
| Thickness |
100 |
100 |
| Window, height x
width |
70 x 78.5 |
100 x 100 |
| Core weight |
80 |
100 |
| Winding weight,
common and series |
6.3 |
100 |
| Tap winding
weight |
100 |
100 |
| Cooling of core
and tap winding |
Forced N2 gas |
ONAN/OFAF |
| Cooling of
common and series winding |
Liquid N2 (with
refrigeration plant) |
ONAN/OFAF |
| Guaranteed %
reactance |
20 |
20 |
| B in core,T |
1.67 |
1.67 |
| J rated(average
of C and S),RMS,A/mm2 |
38 |
2.83 |
| Rated loss,
total |
23 |
100 |
| Overload
capability |
2 p.u.,many
hours |
1.3 p.u.,6 hr,1.5
p.u.,30 min |
On the
performance side, conductor current density is
greatly increased and total rated power loss is
significantly reduced. Overload capability is
somewhat improved. Tolerance of through fault is
severely restricted and is has been seen that the
medium term prospects for overcoming this without
onerous increases in winding AC losses and
conductor cost are, at present, speculation. The
grid transformers are installed in parallel
pairs, in order to improve system security; there
may be an argument for allowing one of the pair
to be superconducting. This is on the basis that,
if one of the transformers has to be disconnected
after a through-fault for up to 1 hr before it
can be reconnected, the statistical reduction in
system security may be tolerable. Table below
shows estimates of savings/expenditure components
based on continuous operation in rated conditions.
It is clear
that expenditure on extra equipments and
materials are offset by the enormous value of the
saved losses (taken over a 10 year period and
discounted at 9.5% per year). However because of
the redundancy built into the system for the
security, the load factor of a grid transformer
is remarkably low and may be taken as 0.225
average and 0.26 RMS. For such load conditions, a
pessimistic estimate suggests that the total
equivalent first cost saving may now become a net
increase of first cost equivalent expenditure of
about 20%.
Table: Cost
savings on continuous full load
| Savings/(expenditure)
% |
| Savings on core
plate |
1 |
| Savings on
continuously transposed Cu |
7 |
| Saving on Cu
losses |
65 |
| Cost of
refrigeration plant |
-21 |
| First cost
equivalent expenditure on refrigerator
drive power |
-6 |
| Total equivalent
first cost saving |
36 |
All values
referred to the overall first cost of the
conventional transformer , which is taken as 100%.
On the other
hand, current UK practice is to have 2
transformers fully rated, normally connected in
parallel. It is thus worth considering an
arrangement of an HTS transformer normally
connecter in parallel with a conventional
transformer normally disconnected but capable of
being switched on quickly when required (e.g.
during a through-fault). Hence saving on the
losses of the conventional transformer will be
very significant.
EXISTING
WORK ON SUPERCONDUCTING TRANSFORMERS:
This
superconducting technology in power transformers
has encouraged several groups to build large
transformers with HTS windings. ABB has built
three phase 630 kVA 18.7/.42 kV transformer using
BiSCCO windings cooled and insulated by liquid
nitrogen at 77 K, and have connected it to the
grid. A Japanese team has built a 500 kVA 6.6/3.3
kV single phase transformer also using BiSCCO
windings cooled and insulated by liquid nitrogen
at 77 K and at 66 K. an American team led by
Waukesha Electric Systems has taken a different
approach and is building a single phase 1 MVA 13.8/6.9
kV transformer, using the cheaper BiSCCO windings
and cooling them to 20-35 K using a cryocooler.
Recently a step
up and step down coreless superconducting
transformer have also been developed in Japan.
It has been
found that the coreless superconducting
transformer has the potential to offer many
advantages, namely reduced power loss; weight
reduction; application as a fault current
limiter; no necessity of tertiary winding at the
receiving side of the transformer and for the
substitution of the shunt reactor if considered
in the total power system. All these advantages
reduce the total life cycle cost compared to that
of conventional transformer.
CONCLUSION:
An HTS
transformer will be substantially cheaper in
first cost, less complex and more reliable in
operation than an LTS transformer. The
requirement for supply power to the refrigeration
plant is much lower per KW of removed heat and
cost saving achieved by eliminating ohmic losses
from the windings is not substantially re-expended
as refrigerator supply power.
The most
difficult technical design problem for a HTS
transformer concerns the through-fault conditions.
This design assuming superconductor properties
cannot recover without disconnection after a
through-fault of even medium size; it can,
however, survive a maximum through-fault, as long
as disconnection is achieved in 170 ms. For grid
transformers, it is worth considering an
operating mode in which HTS and conventional
transformers work in parallel pairs; with the HTS
transformer normally connected and other normally
disconnected but capable of being quickly
connected when required. This would replace the
present UK arrangement of parallel pairs of
conventional transformers (each fully rated),
both being normally connected. The proposal would
avoid the need for recovery capability in the HTS
transformer and also maximize savings on the
power loss in normal operation.
There are
advantages for a grid transformer in having a
conventional design of tap winding, outside the
cryostat because of the constructional
complexities and total heat leak associated with
an in-cryostat design. With oil-less design, the
core losses are removed by gas cooling.
A coreless
superconducting transformer has also been
developed which incorporates many advantages
which reduces the overall total cost.
REFERENCES
- Sykulski,J.K.,Beduz,C.,Stoll,R.L.,Harris,M.R.,Goddard,
K.F.,Yang,Y.,IEE Proc. Power Appl. Jan
1999.
- Okubo,H.,Hikita,M.,Goshima,H.,Sakakibara,H.,Hayakawa,N.;
"High voltage insulation performance
of cryogenic liquids for superconducting
power apparatus",IEEE Trans.Power
Delivery.,1996.
- "Superconducting
transformers" Electrical India Sep.
2001.
- Giese,
R.F.,and Runde,M.; " Assessment
study of superconducting fault current
limiters operating at 77K", IEEE
Trans. Power Delivery., 1993.
- Yamamoto,
M.,Mizukami, N., Ishigohka,T.and
Ohshima,K; "A feasibility study on
superconducting transformer",IEEE
Trans. Magnetics, 1987.
- Riemersma,H.,Barton,M.L,
Litz, D.C, Eckels, P.W, Murphy,J.H and
Roach, J.F; "Application of
superconducting technology to power
transformer",IEEE Trans.Power
Apparatus System.1981.
- Wilkinson,K.J.R
; "Superconducting windings in power
transformer" Proc.IEE,1963.
|