Super-GM Program

Successful tests in Japan and the United States of ac superconducting synchronous generators, coupled with their predicted economic and performance advantages, convinced the Japanese government to launch a major national program in September of 1987 to develop technologies for applying SC to electric power apparatus. This program, called Super-GM (Engineering Research Association Project for Superconductive Generation Equipment and Materials), was specifically chartered to position the Japanese industry into a lead market position for advanced ac synchronous generators using superconducting windings. The initial goal for Super-GM was the design, construction, and test verification of three types of 70 MW-class superconducting generator model machine for establishing technologies to design and manufacture a 200 MW-class pilot machine suitable for commercialization. An additional goal was the parallel development of associated LTS and HTS conductors and related technologies such as refrigeration (Ageta 1996). It is important to note that the Super-GM program was not constrained to use HTS materials. The HTS development under Super-GM was also not specifically directed at an eventual retrofit requirement for the SC field winding. Completely different power applications were considered under the HTS activity: power cable, fault-current limiter, and power leads.

The Super-GM program is administered by the New Energy and Industrial Technology Development Organization (NEDO) as part of the New Sunshine Program of the Agency of Industrial Science and Technology (AIST) under the Ministry of International Trade and Industry (MITI). Super-GM, as a true national program, involves 16 member organizations with representation from the electric utilities; manufacturers of electric power equipment; companies involved in both LTS and HTS research and manufacture of wire and tape; refrigeration and cryogenic suppliers; and independent research institutes such as the Central Research Institute of the Electric Power Industry (CRIEPI). Additional support on collaborative research and consulting is provided by universities and national research organizations such as MITI's Electrotechnical Laboratory (ETL), which has provided basic research and technology assessment. CRIEPI has also assisted Super-GM with benefit and system analysis and electrical analysis for the stator and rotor. NEDO provides direct funding to the Super-GM organization, located in Osaka, in support of the generator design, construction, and test; the conductor research on both LTS and HTS; the refrigeration system; and total system integration and testing. Super-GM is also exploring other electric power devices beside the generator as part of the overall materials effort.

The 1996 budget for Super-GM was approximately $26 million, down from a peak in 1995 of $39 million. The manpower for the total effort averaged approximately 250 people/year in the preceding five years. Nearly $254 million was funded by AIST/MITI from 1988-1996. This figure has been complemented by additional support from the member companies in what amounts to cost-sharing ranging from 20-50% of the contracted amount. The salaries for the technical staff assigned to the Super-GM organization in Osaka, for the most part, are directly paid by the individual companies. Most staff members are rotated from the member companies on a two- to three-year basis.

The Super-GM program entered in 1996 the final installation, testing, and verification phase for the 70 MW-class superconducting generator model machine development. As of April 1997, the machine was undergoing adjustment at the verification test facility prior to verification test; cool-down of the SC rotor with low temperature He gas and liquid helium was to be initiated shortly thereafter. The test facility for the model machine, constructed at Kansai Electric's Osaka power station, is pictured in Fig. 2.1 and shown schematically in Fig. 2.2. Construction, started in June 1994, is now complete. The first rotor (designated "slow response excitation type A") and the "common" stator, both constructed by Hitachi, were delivered and installed in early 1997, with plans for five months of testing starting in June 1997. Following the Hitachi rotor, the Mitsubishi Electric rotor ("slow response B") and the final ("quick response") rotor built by Toshiba will be installed and tested. Testing of the three rotors is scheduled through 1998. The Nb-Ti conductor for these three rotors was supplied, respectively, by Hitachi Cable (for the slow response excitation type A), Sumitomo Electric (for the slow response B), and Furukawa Electric (for the quick response).

Fig. 2.1. External view of the Super-GM test facility at Osaka Power Station.

Fig. 2.2. Super-GM superconducting generator testing schematic.

As shown schematically in Fig. 2.2, the Osaka verification test facility will use a back-to-back motor-generator (M-G) test method with the addition of an induction motor to bring the M-G system up to synchronous speed, compensate for the combined M-G losses, and maintain synchronous speed. The helium refrigeration system, constructed by Mayekawa Manufacturing, is a 100 l/hour closed cycle turbine-expander, screw-compressor system. External liquid nitrogen is used for additional precooling in the high temperature cold box heat exchanger. A preconditioning liquid helium Dewar is used as a buffer to supply liquid to the generator. The liquid helium is introduced into the rotor by means of a helium transfer coupling. The first rotor and stator from Hitachi are undergoing adjustment at the test site. The final testing of the second rotor has been completed at the Kobe works of Mitsubishi Electric. The testing of the first rotor at Hitachi was reported at the August 26-30, 1996, Applied Superconductivity Conference in Pittsburgh, Pennsylvania (Yamaguchi, Takahashi, and Shiobara 1996). These rotating test results indicate that the cooling for the saddle pole SC field winding should be cryostable, due to the supercritical state for the helium, obtained under rotation, combined with the open winding design. The field winding is also well constrained, due to use of a slotted winding, conventional wedges on the straight sections, and retaining rings in the end turns. The cryostable design, along with the good mechanical positioning and the use of both cold and warm electrothermal damper shields, should show high transient capability for the generator during full load testing.

Generator Benefits and Market Projections for Super-GM

The 70 MW-class Super-GM machine is considered a "model" to establish technologies for future design and manufacture of a 200 MW-class "pilot" generator. Provided the 70 MW test is completely successful, a future program targeted at 200-300 MW would be considered. It is anticipated that the costs of an SC and a conventional machine would be nearly identical at a 200-300 MW rating, using a 20-year lifetime. At the time of the WTEC visit, Super-GM managers felt that the major advantage of SC generators was related to "performance advantages for the power system," specifically with respect to steady state and transient stability, improved reactive power capability, and improved ability to tolerate negative sequence fields. Also, Super-GM researchers estimated that SC generators would lead to a ~30% increase in the power transfer limits of the transmission system. Additional economic benefits were an expected 0.5-1% increase in efficiency and a ~50% reduction in size and weight. Use of SC generators would also provide environmental advantages due to reduced oil consumption and reduction of CO2 emissions. The Super-GM market forecast for SC generators, with introduction expected by 2005, using only the national market for Japan, was estimated at $440 million/year with 20-30 units in the 200-600 MW range and two units at 1 GW.

Super-GM LTS and HTS Wire and Tape Research

Since 1988, Super-GM research on superconducting wire and tape in support of power apparatus development has been conducted in parallel with generator research on both LTS and HTS conductors. The LTS effort has primarily focused on development of low ac loss conductors for three different applications: an armature winding manufactured by Furukawa Electric; a shunt reactor manufactured by Sumitomo Electric; and a fault-current limiter manufactured by Hitachi Cable. Three different types of Nb-Ti stranded wires with ~0.1 Ám filaments and a Cu-Ni matrix have been developed by the manufacturers for their respective applications. The conductors have showed reduced ac losses with an acceptable current-carrying capacity of 2-3 kA.

Approaches for obtaining low ac loss Nb3Sn have also been followed, involving 6 different manufacturing processes:

  1. Bronze Process (Furukawa)
  2. Internal Tin (Sumitomo Electric)
  3. In-Situ (Fujikura)
  4. Tube (Showa Wire and Cable)
  5. External Diffusion (Hitachi Cable)
  6. Powder Metallurgy (Kobe Steel)

The Nb3Sn conductors, in general, have not proved to be as good as the Nb-Ti conductors, having nearly one order of magnitude higher hysteresis loss than the Nb-Ti. The transport current is also lower than Nb-Ti at 1-2 kA. The normalized ac quench current at 50 Hz under a dc magnetic field of 0.5 T is roughly equivalent at 200-500 Arms/mm2 for both the Nb-Ti and the Nb3Sn conductors. The Nb-Ti in general has showed a lower ac loss than the Nb3Sn at 50 Hz, ▒ 0.5 T. The Super-GM HTS research on oxide wire and tape development is oriented primarily at a future ability to apply HTS conductor to power applications, including SC generators. The realization of an HTS conductor would simplify the cryogenic design. The basic design approach for an HTS rotor would be nearly identical to an LTS design.

The HTS wire effort, which also started in 1988, was at the time of the WTEC visit making progress using 6 approaches (Yoshida et al. 1996; Chiba et al. 1996):

1./2. Progress in Long Wire Research 3. Progress in Large Current Research 4. Progress in High Current Density Research 5./6. Progress in Large Area Deposition Research for Fault-Current Limiters
Published: September 1997; WTEC Hyper-Librarian