The early activity to demonstrate high field superconducting magnets, which achieved field levels greater than 10 T in a relatively short period, was essentially paralleled by a comparable effort toward applying these Type II superconductors to a wide range of electric power applications. The superconducting power applications can be divided into two categories:
high field > 1 T applications -- generators, motors, fusion, magnetohydrodynamics (MHD), energy storage
The distribution-level FCLs are, for the most part, low field designs, < 1 T, with some of the subtransmission-and transmission-level FCL concepts in the > 1 T range. Some of the low field applications, such as transformers and transmission cables, expose the superconductor to a high level of ac line conditions since they are directly inserted into the ac system; they thus require a conductor design that shows an acceptable ac loss. Some of the FCL designs also expose the superconductor to a high ac field component and therefore need a low-ac-loss conductor. As a result of conductor limitations, work on most "full ac" applications was deferred until the 1980s, when the theory and an ac conductor were more sufficiently developed. The ac transmission cable is an exception as a result of the very low fields of Hop < 0.1 T and the ability to use available Nb3Sn tape conductor, which allows the major field component to be parallel to the plane of the tape. Table 2.1 highlights some of the major electric power components that were constructed and successfully tested during the 1970s and 1980s using liquid helium LTS technology. It is clear from this table and the references it cites that the major world efforts were concentrated in the United States and Japan.
Highlights For Superconducting Electric Power Components
Constructed and Tested from 1970-1990 Using Low Temperature Superconductors
One of the first high field superconducting power applications considered, which this author believes to one of the most important, was to apply high-current-density superconducting wires to electric power synchronous generators. The following discussion on ac synchronous generators also applies to ac synchronous motors, which use an identical design approach, as discussed later for the Reliance Electric motor program. The early machine activity on motors, however, was primarily directed at the dc homopolar design, which allows the use of a "stationary" dc superconducting solenoid (Smith, Kirtley, and Thullen 1975; Appleton 1975). The later activity was concentrated in the United States under U.S. Navy sponsorship and in England.
The basic principal for all rotating electric machines is associated with Faraday's law, which describes the electromagnetic energy conversion related to mechanical movement. In a rotating machine, time-varying voltages are produced in an interconnected set of coils called the "armature winding," which is mechanically moved through a magnetic field, or a magnetic field is mechanically moved past the winding. This magnetic field is produced by the excitation coils or "field winding," which is essentially a dipole magnet for a 2-pole design. If the armature is stationary, it is referred to as the "stator"; the rotating field winding is the "rotor."
It is important to note that the early rationale for SC ac machines primarily emphasized the use of superconductors to achieve higher current densities, which allowed an overall reduction in cross-section and field winding volume compared to ordinary copper-wound rotors. The reduced winding volume thus led to a reduction in the size and weight of the entire machine. The preferred design approach for an ac SC machine developed and demonstrated during the 1970s was to use a stationary room-temperature armature with a rotating SC field winding, which posed difficult problems concerning transferring cryogen into a rotating vacuum-insulated container (Edmonds 1979, 673).
In a synchronous machine, alternating current is supplied to the armature to provide a flux component that rotates in synchronism with the flux component produced by the rotating field winding. The rotor is thus phase-locked at the synchronous speed and under balanced load conditions will see essentially a dc field from the armature. The armature, however, is connected to the electric power system, which experiences load-related electrical disturbances under steady state and transient conditions. These electrical disturbances are reflected back into the armature and produce nonsynchronous ac effects that impact the rotor. Rapid changes in the dc excitation (field forcing) also occur as a result of load changes. The primary armature disturbance is caused by unbalanced loads, which gives rise to negative sequence currents, proportional to the load and degree of unbalance, that counter-rotate at twice the synchronous speed.
Transient events caused by system faults provide the other major source for nonsynchronous ac effects on the rotor. These time-varying fields in the armature, at frequencies different from the synchronous frequency, induce compensating currents to flow in the rotor and produce heating in the dc superconducting winding and structural support. This ac influence, however, can effectively be minimized by incorporating warm and/or cold electromagnetic shields between the two windings that attenuate these nonsynchronous ac fields. The warm shield also provides damping for the mechanical oscillations of the rotor related to the phase of the system. Because the thermal margin for LTS conductors, assuming liquid helium cooling, is between 6 K for Nb-Ti and 14 K for Nb3Sn, the shielding must be carefully designed to minimize ac heating of the field winding and to prevent degradation of the superconductor Jc and magnetic field capability and possible normalization during extreme transient conditions.
Following the first demonstration of an ac synchronous machine with a rotating SC field winding in 1971, major research programs were initiated in the United States, Europe, Japan, and the USSR -- the United States and Japan being the major players (Thullen et al. 1971). During the 1970s, a number of machines were built and successfully tested, highlighted in Table 2.1, that completely demonstrated that ac superconducting machines could be built in large sizes suitable for electric utility installation. The 12,000 rpm SC four-pole rotor test by Westinghouse for the United States Air Force (USAF) Wright-Patterson Laboratories, demonstrated, due to the high centrifugal loads on the superconducting winding, that larger machine diameters with ratings near 1,000 MVA could be successfully constructed and operated with liquid helium (Parker et al. 1975).
The principal arguments advanced during this period were that ac superconducting machine technology could achieve (1) efficiency improvements near 1%, (2) decreased size and weight for equivalent ratings, (3) ability to manufacture larger size generators than is possible with conventional technology, (4) improved steady state and transient system performance, and (5) reduced life-cycle costs, assuming reliability and maintenance comparable to existing units.
The prospect of approximately 1% increased efficiency for the SC machine offered to the utilities substantial savings in annual operating costs as a result of reduced fuel consumption. The savings in fuel costs were, in fact, so large over the ~40-year lifetime of conventional machines that they could almost completely offset the initial cost of the generator. This savings, however, was completely dependent on the SC generator having a reliability profile identical to that of a conventional unit. Furthermore, if the SC generator experienced even one additional day of outage per year compared to a conventional unit, the efficiency-derived savings would be essentially eliminated. The reliability and maintenance profile for an SC generator must therefore be identical with a conventional machine to ensure its economic benefit. Because it would take many years to produce sufficient operating experience to obtain an acceptable reliability profile, the utilities adopted a reserved attitude towards the projected savings resulting from improved efficiency.
SC ac machine technology, however, in addition to efficiency improvement, offered an impressive list of system improvements that caught the interest of and excited utility customers. Studies by the Japanese under their Super-GM program have confirmed and cited these system improvements as a major driver for the eventual commercialization of superconducting generators (Ogawa 1992).
To summarize the commercial interest in developing superconducting generators, technical interest was originally driven by the ability of SC generators to increase current density, which permitted higher magnetic fields and allowed a reduction in weight and size. In addition, the technology made possible the realization of increased machine efficiency because of the elimination of I2R heating in the field winding, which quickly became the focal attraction of SC machines. Subsequent experience revealed marked improvement in SC generator system interactions over those of conventional machines, and this aspect turned out to be particularly attractive to utility customers. The cryogenic aspects, although of concern, were not viewed as a limitation for utility consideration of LTS ac synchronous machines. The utilities, however, did view with great suspicion the overall operational implications of cryogenics, that is, the unlikely prospect that a generator using liquid helium could be constructed and operated with reliability and maintenance profiles identical to those of conventional generators. They viewed the added requirement for a refrigerator more as a reliability issue than as a cost or machine efficiency issue.
The U.S. effort on ac machines was essentially terminated in 1981 with the cancellation of the Westinghouse-EPRI (Electric Power Research Institute) 300 MVA program. The efforts on almost all of the SC generator and other power-related programs were also markedly curtailed at that time, which was largely a result of the decrease in interest and funding from the federal government and the unwillingness of industry to go forward on its own. Companies worldwide, except for those in Japan, almost collectively decided to reduce their effort and await further developments and an essential improvement in the marketplace created by increased electric power demand.
Technical development and demonstration for almost all of the various SC power applications was thus fairly well in place towards the end of the 1980s. Market conditions in the electric power sector, however, were fairly soft worldwide, which severely restricted the industry's ability to invest in new technology. Reluctance to invest in R&D, which was common in the United States, was not completely universal, and selected power-related development activities in SC continued to flourish. The Japanese officially launched their Super-GM program in September 1987, with a long-range objective of developing superconducting generators and other electric power applications that they expected to offer for sale to the utility market following the turn of the century (Ogawa 1992).
It was during this time period that the discovery was made of high temperature superconductors (HTS), which offered the advantage of cooling via liquid nitrogen instead of liquid helium. In the United States there was almost an immediate resurrection of interest in superconducting applications, with the Department of Energy (DOE) and Defense Advanced Research Projects Agency (DARPA) taking the lead in research and development of electric power applications. In 1988, DOE began its Superconductivity Program for Electric Power Systems, which primarily supported work at the national laboratories focused on development of wire and tape HTS materials for use in electric power equipment. This DOE program has evolved into the present effort, established in 1993, called the Superconductivity Partnership Initiative (SPI), which is helping to fund the industrial development of electric power components using HTS materials. The DOE SPI program initiated four industry-led projects directed at development of key superconducting electric power applications:
These four projects also teamed with one or the other of the two major U.S. HTS wire and tape manufacturers, American Superconductor and Intermagnetics General, and with a utility that represented an end user. DOE recently awarded Phase II programs to develop "precommercial" prototypes to Reliance Electric for a 5,000 hp motor and to Lockheed Martin for a 15 kV-class fault-current limiter.
The advantages offered by HTS wire or tape over conventional LTS materials that rely on liquid helium may not be completely obvious. For some applications, the overall impact on efficiency of HTS technology due to operation at liquid nitrogen temperature may be insignificant in comparison to LTS technology operating at liquid helium temperature. For example, for a large (>100 MVA) ac synchronous machine, the impact on the machine efficiency derived from the 25-50 times reduction in refrigerator power consumption offers no major economic advantage; even complete elimination of refrigerator power consumption would only show an improvement in machine efficiency of ~0.02% for a 300 MVA rating (Blaugher 1996). Of more importance than the efficiency improvement is that use of a liquid nitrogen ambient would lead to reduced capital costs for the refrigeration plant and reduce the complexity of the cryogenic design. Even more important would be the projected improvement in the entire cryogenic system with respect to reliability.
The performance of an HTS superconducting coil in a liquid nitrogen environment would show unparalleled stability compared to LTS performance. If copper is assumed to be the stabilizer for an HTS conductor, the resulting specific heat for the conductor would be approximately three orders of magnitude higher than for a conventional 4.2 K copper-stabilized LTS conductor. Thus, the HTS conductor would be inherently more stable, even with a possible reduction in current sharing caused by the increase in resistivity for the copper matrix. In addition, the critical heat flux, i.e., the peak nucleate boiling or transfer from nucleate to film boiling, is higher for 77 K liquid nitrogen than for liquid helium by over an order of magnitude. These combined thermal properties would further enhance the operational stability for an oxide coil over comparable LTS construction. It would thus be possible to maintain adiabatic stability at filament sizes several orders of magnitude larger than those of LTS conductors. The ability of HTS conductors to tolerate conductor movement due to Lorentz forces would also be improved over LTS conductors (Blaugher 1996).
These factors also contribute to the HTS coils being more tolerant of ac loss. Increased ac loss tolerance follows as a result of the ability to tolerate larger temperature excursions during ac transient events, which would be mostly applicable to the ac synchronous generators and motors. Exposure to steady state ac currents, as experienced for the transmission line and transformer, would continue to demand a low-ac-loss conductor to minimize the heat load and maximize the operating efficiency. The small efficiency improvement of 0.1-0.2% for these applications would be negated if high ac losses were exhibited.
Table 2.2 compares various electric power applications with regards to the HTS wire or tape performance requirements for prototyping and eventual commercialization, as defined by the industry. (The current, field, and mechanical performance requirements could apply equally to LTS conductors.) The operating temperatures listed apply only to HTS wire; LTS coils normally operate at 4.2-8 K.
The last column of Table 2.2 lists the cost target for HTS conductors in $/kAm. Cost considerations follow directly from the earlier observation that an SC coil is cost-effective only if the amp-turns for a given coil or cable show a marked advantage over ordinary copper conductor. At present, the cost for conventional LTS Nb-Ti wire is ~$2/kAm; Nb3Sn costs two to five times more, depending on whether the conductor is multifilament or tape. The lowest cost listed in Table 2.2 for HTS systems, $10/kAm, would thus be consistent with the cost of Nb3Sn and is judged by most industry experts to be acceptable for all electric power applications. The present cost estimate for Bi wire, which has far lower performance levels than Nb3Sn, is at least 1 to 2 orders of magnitude higher than the cost estimate for Nb3Sn. The present high cost of HTS conductors is not likely, however, to limit attempts to construct and demonstrate the various applications, which is born out by recent successful tests of an HTS transmission line (Scudiere et al. 1996), HTS fault-current limiter (Leung et al. 1996), and HTS synchronous motor (Schiferl, Zhang, Shoykhet et al. 1996; Schiferl, Zhang, Driscoll et al. 1996). The commercialization, i.e., market price, for all of these applications demands that the lowest price be offered in order for the SC components to be cost-competitive with conventional nonsuperconducting devices (Bray 1995).
High Temperature Superconducting Applications: Industry's Wire Performance Requirements
a. Current density for individual high temperature superconducting
b. Current for individual wire; distribution cables, with multiple wires, require current near 10 kA
c. Cable requirement
Table 2.3 compares the key parameters that utilities will use when evaluating the purchase of an SC power device against existing or alternative technologies. This table compares all of the current SC power-related applications with respect to system performance, reliability and maintenance, efficiency, operating lifetime, and installed cost against representative competing technologies.
Although each individual utility will approach its needs differently and conduct its evaluation of a given component according to different priorities, and other unlisted parameters such as environmental impact may also guide decisionmaking; nevertheless, certain concerns appear to be significant for all utilities. The electric utilities are extremely sensitive to cost, and stated life-cycle cost is one of the most important elements in their evaluation. Operational requirements must also be consistent with their normal way of doing things, i.e., servicing and maintenance procedures should be similar to their normal practice. Coupled with this last requirement is the reliability and maintenance cost factor for the component, which must be comparable to that of their present equipment. The utilities must be able to integrate and service these new SC components, with some modest additional training, using existing power station or utility personnel. The maintenance cycles must also be comparable with standard utility practice. Advertised efficiency improvements, although universally desirable, are admittedly uncertain, due to insufficient operating experience; thus, the industry perception is that promoters cannot accurately predict a reliability and maintenance profile.
The SC or cryogenic applications should, in principle, offer a longer operating lifetime than their conventional counterparts. All of the conventional technologies commonly experience degradation to the insulation due to thermal aging, which should be insignificant for the cryogenic applications. The insulation in an SC rotor for a generator or motor should not degrade; hence, the usual rewinding maintenance should not be required during its 30-40 year operating lifetime.
Comparison of Superconducting Electric Power Applications to Conventional Technologies
1. Includes unit cost, siting, and system support, i.e.,
refrigeration, power conditioning, etc., compared to conventional
2. May require additional components, i.e., circuit breakers and/or current limiters
3. Requires low ac loss conductor
In the absence of thermal aging, life-cycle costs for the generator or motor improve accordingly. The life-cycle costs, as currently applied by the utilities, include the total capital and installation cost, all operating costs, all known maintenance costs such as rotor and armature rewinds, and annual inspection and refurbishing; the total is then depreciated over a 30- or 40-year lifetime. The overall plant cost, i.e., for the construction of the power plant, is also rolled up with the electric power components (generators, transformers, switch gear, etc.) to arrive at an overall life-cycle cost for the entire facility. As stated before, the success for SC applications is highly dependent on their ability to show life-cycle costs equal to or better than those of conventional components.
Utility people generally accept the predictions of higher efficiency and lower life-cycle costs for SC technology and completely understand their interdependence with reliability and maintenance. For this reason, system advantages over conventional components are key to obtaining utility interest in a given SC device. If the system performance offers enough of an incentive, utility staff will use this feature to justify the initial purchase cost and also explain to senior management the prospect for improved efficiency and lower life-cycle costs. The present tight fiscal climate as a result of deregulation may weaken the life-cycle argument, reinforcing the primary importance of the initial purchase price. A high-cost SC component thus may have a disadvantage compared to a conventional product.
Superconducting synchronous generators, underground transmission, and fault-current limiters all appear to offer unique system advantages along with reduced life-cycle costs. Superconducting underground transmission provides improved impedance-matching compared with conventional underground cable, which normally requires series/shunt compensation for lengths greater than 20 miles (Forsyth 1983, 285). It is not by accident that all of these highly favorable electric power components are currently under development by the DOE SPI program and Japan's MITI-sponsored SC programs.