As a very attractive application of superconductivity, Superconducting Magnetic Energy Storage (SMES) has been studied for 30 years, and with the development of superconductor technology, it has now just come to the stage of construction for practical SMES systems. The first systematic study by R. W. Boom aimed at having a very large-scale SMES for diurnal load leveling in electric power line systems were followed by many other design studies and projects. The representative projects at present are the USA ETM 20MWh project and 500kWh one by Babcock & Wilcox. The former is the first project concerning a practical scale system, which, however, had to be closed in 1982 before construction. The latter is now in progress and is to be installed in Anchorage ML & P in 1997 as the first commercial SMES system. The main points of SMES design are economic design of coils, support structures for magnetic force, and power conditioning systems. The basic idea for these are explained first. Seen from the side of application, it should be characterized by the stored energy (W) and the power (P) from it, and wide variety of applications ranging from SMES's for transmission line stabilization to those for diurnal load leveling can be discussed in P-W map. The practical region and what scaling steps in the development have taken place are shown in the map. Next, points of magnet technology varies with the scale of SMES. They are characterized by the average current density (Jav) and W, and are discussed in Jav-W map which has an allowable domain restricted by some limitations. The above two maps may help to figure out the whole feature of SMES systems. Finally, the present status of SMES development and problems are stated with regard to ETM design, B & W project and Micro-SMES, as well as on the state of the art of Japanese SMES research, in some detail.
To compare stability in AC use of NbTi and Nb3Sn superconductors, a critical quench margin of the superconductors has been defined. The comparison has been done based on the critical quench margin obtained by numerical calculations on conditions of AC operation in the AC field. The comparison has also been done on other conditions of DC operation in DC fields superposed by AC fields. Through these comparisons, the advantages of Nb3Sn superconductors over NbTi ones have been found as follows: As for AC operation in AC field (commercial frequency, amplitude: 0.1-0.5T) in power apparatuses, such as fault current limiters and shunt reactors, the high current density type Nb3Sn (Jc: around 4, 000A/mm2 at 0.5T, total loss: around 20kJ/m3/cycle in a peak field of 0.5T, 50Hz) is more stable than NbTi. As for AC operation in an AC field (commercial frequency, amplitude: 0.5-1.5T) in power apparatuses, such as armature windings of superconducting generators and power transformers, the low AC loss type Nb3Sn (Jc: around 2, 000A/mm2 at 0.5T, total loss: around 2kJ/m3/cycle in a peak field of 0.5T, 50Hz) is more stable than NbTi. Furthermore, as for DC operation in a DC field (5T) superposed by AC field (amplitude: peak field of 0.1-1.5T, frequency: 2.5-350Hz), the high current density tpe Nb3Sn is more stable than NbTi.
The per-cycle AC loss properties at 77K have been investigated on the (Bi, Pb)2Sr2Ca2Cu3Ox superconducting rod-form wire with a ceramics cylinder embedded by Ag sheaths through two kinds of measuring methods. One is the transport method measuring the resistive voltage under AC transport currents and another is the magnetic method measuring the magnetization curve under AC magnetic fields. The loss values for the former are 20-50% smaller than those for the latter. This difference is explained by the anisotropy in critical current densities Jc due to a current direction that the Jc value for a longitudinal direction giving rise to the AC transport losses is larger than that for a circumferential direction to the AC magnetic losses. Apart from the difference in magnitude, the transport and magnetic losses increase with increasing frequency f at a fixed magnetic field B0 (0.2mT≤B0≤45mT) except for around the field Bp (≈5mT) for full flux-penetration, where the losses are nearly independent of f. Numerical calculations based on the critical state model show that the f dependence of the losses is caused by the losses in Ag sheath. As B0 increases, the f-dependent term due to the losses in the Ag sheath increases linearly with B02 in fields B0<<Bp, deviates from the behavior with an upward curvature at around Bp, and falls again on the B02 dependence in fields B0>>Bp. This behavior comes from the fact that the electric field in the Ag sheath is produced by magnetic flux in both the Ag sheath and the ceramic core.
Effects of pre-loading at room temperature on critical current at 4.2K of multifilamentary Nb3Sn superconducting composite wire heat-treated at 973K for 43 (sample A) and 259ks (B) were investigated. The main reults are summarized as follows. (1) It was predicted, based on Ekin's scaling law in combination with the elastic/plastic mechanical calculation, that the upper critical magnetic field and critical current can be improved by the pre-loading treatment through the control of the residual strain of Nb3Sn. This prediction was verified experimentally. (2) The strength distribution of Nb3Sn in sample A was estimated by applying the Weibull distribution function. The shape and scale parameters were 12 and 0.80GPa (standard length 1m), respectively. From these results, the average strengths for the length of 3 and 300m were calculated to be 0.70 and 0.48GPa, being lower than the strength of 1.0GPa for the short samples with a length of 25mm. (3) Using the strength distribution of Nb3Sn, Ekin's scaling law and the experimental data for the short samples, an attempt was made to deduce the influence of pre-loading on the critical current for long samples. It was shown for sample A with a length of 300m that the pre-stress level on the composites, at which the critical current is reduced due to the breakage of Nb3Sn, would be reduced by 20% in comparison with that for short samples of 25mm.