SUPERCONDUCTING TRANSFORMERS -- OVERVIEW

Transformers represent one of the oldest and most mature elements in a power transmission and distribution network. From the point of electricity generation at a power plant, where extremely high voltages are needed to "push" large amounts of power into the grid, to the end user of electricity in a home or office, where typical appliances operate at much lower voltages (100-200 volts), transformers are needed to effect voltage conversions. At each conversion point, energy is lost, primarily in the form of wasted heat from changing electrical and magnetic fields in the copper (coil), iron (core), tank, and supporting structure. Even when the transformer is "idling," so-called "no-load losses" (NLL) are generated in the core. Research over the last 50 years has succeeded in reducing NLL by a factor of three while increasing core costs by a factor of two. Recent substitution in distribution transformers (ratings below about 100 kVA) of amorphous metals for silicon iron core material has reduced NLL further, but this material has not been used in the cores of power transformers (ratings greater than 500 kVA). When a transformer is under a loaded condition, Joule heating (I²R losses) of the copper coil adds considerably to the amount of lost energy. In spite of the fact that today's utility power transformer loses less than 1% of its total rating in wasted energy, any energy saved within this one percent represents a tremendous potential savings over the expected lifetime of the transformer.

In a conventional power transformer, load losses (LL) represent approximately 80% of total losses. Of this load loss, 80% are I²R losses. The remaining 20% consists of stray and eddy current losses. To date, efforts to reduce load losses have been directed toward the latter. Unlike copper and aluminum, superconductors present no resistance to the flow of dc electricity, with the consequence that I²R losses become essentially zero, thereby creating the potential for a dramatic reduction in overall losses. In ac operation, the superconductor in an HTS transformer experiences a type of eddy current loss: both the heat produced by this loss (although extremely small in comparison to the energy lost in conventional materials) and heat conducted into the lower temperature regions of the superconducting transformer need to be removed through refrigeration. Even with the added cost of refrigeration, HTS transformers in the 10 MVA and higher range are projected to be substantially more efficient and less expensive than their conventional counterparts.

Motivation for developing superconducting transformers is not based solely on economic considerations of lowering total owning costs (initial capital cost + capitalized cost of load and no-load losses over the transformer's effective life). With limited new siting availability in urban areas, the anticipated 2% annual growth in power demand means that existing sites must be uprated with higher power capabilities. Many existing sites are indoors or adjacent to buildings, which restricts the use of most oil-filled transformers. The inherent dangers of oil-filled devices are totally eliminated by application of superconducting technology where the only coolant required is benign (nitrogen as opposed to oil). Consequently, superconducting transformers operating either with a refrigerated coil or one cooled with liquid nitrogen pose no fire hazards and no threat to the environment comparable to that posed by leaks of flammable oils and toxic chemicals such as PCBs.

Serious interest in superconducting transformers began in the early 1960s as reliable low temperature superconductors based on Nb-Ti and Nb₃Sn became available. Analysis of the feasibility of such LTS transformers concluded that the high refrigeration loads required to keep the LTS materials at 4.2 K made the LTS transformers uneconomical. A major reduction in refrigeration costs and/or the discovery of materials that superconduct at much higher temperatures would be required to improve the economic attractiveness of these electric power applications. In the mid-1970s Westinghouse conducted an exhaustive design study of a 1,000 MVA, 550/22 kV generator step-up unit; it found that current transfer, overcurrent operation, and protection remained persistent problems.

Since 1980, development of LTS transformers has been conducted primarily by ABB and GEC-Alsthom in Europe and by various utilities, industries, and universities in Japan. Advances in production of long-length ultrafine multifilamentary Nb-Ti conductor and high resistivity Cu-Ni matrix materials have assisted in the reduction of ac losses. Feasibility of weight reduction and higher efficiencies has been demonstrated on smaller devices with ratings smaller than 100 kVA: single-phase 80 kVA (Alsthom), 30 kVA (Toshiba), and a three-phase 40 kVA (Osaka University). Larger units have also been constructed and tested successfully. A single-phase 330 kVA transformer built by ABB included provisions for fault-current limiting and quench protection. Kansai Electric Power Company reported the development of an LTS transformer utilizing Nb₃Sn conductor. One phase of this three-phase 2,000 kVA unit operated at 1,379 kVA without quenching and transferred fault current to parallel coils under quench condition.