Superconducting magnetic energy storage

Superconducting Magnetic Energy Storage (SMES) systems store energy in the magnetic field created by the flow of direct current in a superconducting coil which has been cryogenically cooled to a temperature below its superconducting critical temperature. A typical SMES system includes three parts: superconducting coil, power conditioning system and cryogenically cooled refrigerator. Once the superconducting coil is charged, the current will not decay and the magnetic energy can be stored indefinitely. The stored energy can be released back to the network by discharging the coil. The power conditioning system uses an in erter!rectifier to transform alternating current "A#$ power to direct current or con ert %# back to A# power. The in erter!rectifier accounts for about &'() energy loss in each direction. SMES looses the least amount of electricity in the energy storage process compared to other methods of storing energy. SMES systems are highly efficient* the round'trip efficiency is greater than +,). %ue to the energy re-uirements of refrigeration and the high cost of superconducting wire, SMES is currently used for short duration energy storage. Therefore, SMES is most commonly de oted to impro ing power -uality. .f SMES were to be used for utilities it would be a diurnal storage de ice, charged from baseload power at night and meeting peak loads during the day.

Contents

/ Ad antages o er other energy storage methods & #urrent use ( #alculation of stored energy 0 Solenoid ersus toroid , 1ow'temperature ersus high'temperature superconductors 2 #ost 3 Technical limitations 4 5eferences + Manufacturers /6 See also

Advantages over other energy storage methods

There are se eral reasons for using superconducting magnetic energy storage instead of other energy storage methods. The most important ad antages of SMES is that the time delay during charge and discharge is -uite short. 7ower is a ailable almost instantaneously and ery high power output can be pro ided for a brief period of time. Other energy storage methods, such as pumped hydro or compressed air ha e a

substantial time delay associated with the energy con ersion of stored mechanical energy back into electricity. Thus if a customer8s demand is immediate, SMES is a iable option. Another ad antage is that the loss of power is less than other storage methods because electric currents encounter almost no resistance. Additionally the main parts in a SMES are motionless, which results in high reliability.

Current use
There are se eral small SMES units a ailable for commercial use and se eral larger test bed pro9ects. Se eral / M: units are used for power -uality control in installations around the world, especially to pro ide power -uality at manufacturing plants re-uiring ultra'clean power, such as microchip fabrication facilities. These facilities ha e also been used to pro ide grid stability in distribution systems. SMES is also used in utility applications. .n northern :isconsin, a string of distributed SMES units was deployed to enhance stability of a transmission loop. The transmission line is sub9ect to large, sudden load changes due to the operation of a paper mill, with the potential for uncontrolled fluctuations and oltage collapse. %e elopers of such de ices include American Superconductor. ETM "engineering test model$ is a large SMES with a capacity of appro;imately &6 M:<h, capable of pro iding 066 M: of power for /66 seconds or /6 M: of power for & hours.

Calculation of stored energy

The magnetic energy stored by a coil carrying a current I is gi en by one half of the inductance of the coil times the s-uare of the current.

:here E = energy measured in 9oules L = inductance measured in henries I = current measured in amperes >ow let?s consider a cylindrical coil with conductors of a rectangular cross section. The mean radius of coil is R. a and b are width and depth of the conductor. f is called form function which is different for different shapes of coil. ";i$ and "delta$ are two parameters to characteri@e the dimensions of the coil. :e can therefore write the magnetic energy stored in such a cylindrical coil as shown below. This energy is a function of coil dimensions, number of turns and carrying current.

&

:here E = energy measured in 9oules I = current measured in amperes f(,) = form function, 9oules per ampere'meter N = number of turns of coil

Solenoid versus toroid

Aesides the properties of the wire, the configuration of the coil itself is an important issue from a mechanical engineering aspect. There are three factors which affect the design and the shape of the coil. They are inferior strain tolerance, thermal contraction upon cooling and lorent@ forces in a charged coil. Among them, the stain tolerance is crucial not because of any electrical effect, but because it determines how much structural material is needed to keep the SMES from breaking. Bor small SMES systems, the optimistic alue of 6.() stain tolerance is selected. Toroidal geometry can help to lessen the e;ternal magnetic forces and therefore reduces the si@e of mechanical support needed. Also, due to the low e;ternal magnetic field, toroidal SMES can be located near a utility or customer load. Bor small SMES, solenoids are usually used because they are easy to coil and no pre' compression needed. .n toroidal SMES, the coil is always under the compression by the outer hoops and two disks, one of which is on the top and the other is on the bottom to a oid breakage. #urrently, there is little need for toroidal geometry for small SMES, but as the si@e increases, mechanical forces become more important and the toroidal coil is needed. Many of the older large SMES concepts usually featured a low aspect ratio solenoid appro;imately /66m in diameter buried in earth. At low e;treme of si@e is the concept of micro'SMES solenoids, for energy storage range near / MC.

Low-temperature versus high-temperature superconductors

Dnder steady state conditions and in the superconducting state, the coil resistance is negligible. Eowe er, the refrigerator necessary to keep the superconductor cool re-uires electric power and this refrigeration energy must be considered when e aluating the efficiency of SMES as an energy storage de ice. Although the high'temperature superconductor "ETS#$ has higher critical temperature, flu; lattice melting takes place in moderate magnetic fields around a temperature lower

than this critical temperature. The heat loads that must be remo ed by the cooling system include conduction through the support system, radiation from warmer to colder surfaces, A# losses in the conductor" during charge and discharge$, and losses from the coldFto' warm power leads that connect the cold coil to the power conditioning system. #onduction and radiation losses are minimi@ed by proper design of thermal surfaces. 1ead losses can be minimi@ed by good design of the leads. A# losses depend on the design of the conductor, the duty cycle of the de ice and the power rating. The refrigeration re-uirements for ETS# and low'temperature superconductor "1TS#$ toroidal coils for the baseline temperatures of 33 G, &6 G and 0.& G, increases in that order. The refrigeration re-uirements here is defined as electrical power to operate the refrigeration system. As the stored energy increases by a factor of /66, refrigeration cost only goes up by a factor of &6. Also, the sa ings in refrigeration for an ETS# system is larger the that for an 1TS# system, range from 26) to 36) for these cases.

Cost
Are ETS# systems more economic than 1TS# systemsH .t depends because there are other ma9or components determining the cost of SMES: #onductor consisting of superconductor and copper stabili@er and cold support are ma9or costs in themsel es. They must be 9udged with the o erall efficiency and cost of the de ice. Other components, such as acuum essel, insulation, has been shown to be a small part compared to the large coil cost. The combined costs of conductors, structure and refrigerator for toroidal coils are dominated by the cost of the superconductor. The same trend is true for solenoid coils. ETS# coils cost more than 1TS# coils by a factor of & to 0. :e e;pect to see the cheaper cost for ETS# due to lesser refrigeration re-uirements but this is not the case. So, why is the ETS# system more e;pensi eH To gain some insight consider a breakdown by ma9or components of both ETS# and 1TS# coils corresponding to three typical stored energy le els, &, &6 and &66 M:<h. The conductor cost dominates the three costs for all ETS# cases and is particularly important at small si@es. The principal reason lies in the comparati e current density of 1TS# and ETS# materials. The critical current "Jc$ of ETS# wire is lower than 1TS# wire generally in the operating magnetic field, about , to /6 teslas "T$. Assume the wire costs are the same by weight. Aecause ETS# wire has lower Cc alue than 1TS# wire, it will take much more wire to create the same inductance. Therefore, the cost of wire is much higher than 1TS# wire. Also, as the SMES si@e goes up from & to &6 to &66 M:h, the 1TS# conductor cost also goes up about a factor of /6 at each step. The ETS# conductor cost rises a little slower but is still by far the costliest item. The structure costs of either ETS# or 1TS# go up uniformly "a factor of /6$ with each step from & to &6 to &66 M:<h. Aut ETS# structure cost is higher because the strain tolerance of the ETS# "ceramics cannot carry much tensile load$ is less than 1TS#, such as >bTi, which demands more structure materials. Thus, in the ery large cases, the ETS# cost can not be offset by simply reducing the coil si@e at a higher magnetic field.

Maybe it is worth noting here that the refrigerator cost in all cases is so small that there is ery little percentage sa ings associated with reduced refrigeration demands at high temperature. :hat does this meanH .t means that if a ETS#, AS##O for instance, works better at a low temperature, say &6G, it will certainly be operated there. Bor ery small SMES, the reduced refrigerator cost will ha e a more significant positi e impact. #learly, the olume of superconducting coils increases with the increase of the stored energy. Also, we can see that the 1TS# torus ma;imum diameter is always smaller for a ETS# magnet than 1TS# due to higher magnetic field operation. .n the case of solenoid coils, the height or length is also smaller for ETS# coils, but still much higher than in a toroidal geometry "due to low e;ternal magnetic field$. An increase in peak magnetic field yields a reduction in both olume "higher energy density$ and cost "reduced conductor length$. Smaller olume means higher energy density and cost is reduced due to the decrease of the conductor length. There is an optimum alue of the peak magnetic field, about 3 T in this case. .f the field is increased past the optimum, further olume reductions are possible with minimal increase in cost. The limit to which the field can be increased is usually not economic but physical and it relates to the impossibility of bringing the inner legs of the toroid any closer together and still lea e room for the bucking cylinder. The superconductor material is a key issue for SMES. Superconductor de elopment efforts focus on increasing Cc and strain range and on reducing the wire manufacturing cost

Technical limitations
The energy content of current SMES systems is usually -uite small. Methods to increase the energy stored in SMES often resort to large'scale storage units. As with other superconducting applications, cryogenics are a necessity. A robust mechanical structure is usually re-uired to contain the ery large 1orent@ forces generated by and on the magnet coils. The dominant cost for SMES is the superconductor, followed by the cooling system and the rest of the mechanical structure.

Mechanical support ' >eeded because of lorent@ forces. Size ' To achie e commercially useful le els of storage, around / I:<h "0 TC$, a SMES installation would need a loop of around /66 miles "/26 km$. This is traditionally pictured as a circle, though in practice it could be more like a rounded rectangle. .n either case it would re-uire access to a significant amount of land to house the installation, and to contain the health effects noted below. Manufacturing ' There are two manufacturing issues around SMES. The first is the fabrication of bulk cable suitable to carry the current. Most of the superconducting materials found to date are relati ely delicate ceramics, making it difficult to use established techni-ues to draw e;tended lengths of ,

superconducting wire. Much research has focussed on layer deposit techni-ues, applying a thin film of material onto a stable substrate, but this is currently only suitable for small'scale electrical circuits.

Infrastructure ' The second problem is the infrastructure re-uired for an installation. Dntil room'temperature superconductors are found, the /66 mile "/26 km$ loop of wire would ha e to be contained within a acuum flask of li-uid nitrogen. This in turn would re-uire stable support, most commonly en isioned by burying the installation. ritical current ' .n general power systems look to ma;imi@e the current they are able to handle. This makes any losses due to inefficiences in the system relati ely insignificant. Dnfortunately the superconducting properties of most materials break down as current increases, at a le el known as the critical current. #urrent materials struggle, therefore, to carry sufficient current to make a commercial storage facility economically iable. ritical !agnetic fiel" ' 5elated to critical current, there is a similar limitation to superconducti ity linked to the magnetic field induced in the wire, and this too is a factor at commercial storage le els #ossible $"%erse &ealth effects ' The biggest concern with SMES, beyond possible accidents such as a break in the containment of li-uid nitrogen, is the ery large magnetic fields that would be created by a commercial installation, which would dwarf the magnetic field of the Earth. 1ittle is known about the long term effects of e;posure to such fields, so any installation is likely to re-uire a significant buffer @one around and abo e it to protect humans and wildlife.

eferences

/$ Sheahen, T., 7. "/++0$. .ntroduction to Eigh'Temperature Superconducti ity. 7lenum 7ress, >ew Jork. pp. 22, 32'34, 0&,'0(6, 0(('002. &$ El':akil, M., M. "/+40$. 7owerplant Technology. McIraw'Eill, pp. 24,'24+, 2+/'2+,. ($ :olsky, A., M. "&66&$. The status and prospects for flywheels and SMES that incorporate ETS. 7hysica # (3&F(32, pp. /0+,F/0++. 0$ Energy storage basics and comparisons ,$ #ost Analysis of Energy Storage Systems for Electric Dtility Applications 2$ SMES presentation 3$ 7ower #onditioning SMES unit 2

4$ 1oyola SMES summary +$ 1arge'Scale Energy Storage Systems