This thesis focuses on electromagnetic design of light weight and high power density superconducting synchronous machines for 10 MW class wind turbine generators. There are three motivations of this research like the following; (1) to propose the world's first fully superconducting generator having YBCO field coils and MgB2 armature windings, (2) electromagnetic design for three kinds of superconducting and a permanent magnet type wind turbine generators and (3) investigating the electromagnetic characteristics of 10 MW class direct-drive wind turbine generators by means of generator size, weight, HTS wire length, generator losses and so on.

Wind generation is well known as one of the clean energy and the installation of wind energy is increasing annually in global levels. The size of wind turbines has been also increasing from the economical point of view. There are some proposals for over 10 MW class generators. However, the generator's weight and tower cost increase with the generator capacity. It is important to develop lightweight and high power density generators for large capacity wind turbine generators.

High-temperature superconducting (HTS) technology is one of the key solutions. Superconducting wind turbine generators have a potential for realizing compact and high-power density generators and have been studied actively in the world. In general, these generator structures are made of superconducting field coils and copper armature windings.

Fully superconducting generator which has superconducting field and armature windings has a potential to develop more compact and higher output density generator than other superconducting machines. The two coils are put in the same cryostat and it results in air gap reduction. Finally, it is possible to reduce not only generator size. However, the fully superconducting generator has a technical challenging issue; how to reduce AC loss at the superconducting armature windings. In addition to them, the total HTS wire length will be increase and results in higher cost machine. If multifilament MgB2 superconducting wires are applied into the superconducting generators, it is possible to reduce the AC losses. And also, the MgB2 wire is so lower cost than usual HTS wires YBCO and BSCCO, that the cost issue can be solved.

Chapter 1 shows the introduction of this study. Firstly, current status of global wind energy installation and large scale wind turbine generator systems are described. The next, superconducting wind turbine generators studied by some company and university groups are introduced as a solution of the issues after explaining technical challenging issues for over 10 MW class. Finally, the overview of this thesis is shown.

Chapter 2 refers design conditions for 10 MW class wind turbine generators. Firstly, the initial design conditions such as output power, current voltage and so on are shown. The next, four generator structures and their cooling system design schemes are described. And also, the design concepts for generator components such as HTS field coils, copper or MgB2 armature windings and back iron have been explained.

Chapter 3 is for electromagnetic design of permanent magnet type synchronous generators (PMSG). Electromagnetic design and characteristics of the PMSG were investigated with 2D finite element method analysis (FEM). Two kinds of electromagnetic characteristics such as steady and transient state are analyzed.

Firstly, the electromagnetic characteristics in steady state have been investigated. The results showed that the diameter and weight of 10 MW class PMSG was over 13 m and 230 tons respectively. Especially, 13.1 tons of permanent magnet was included in the weight. The current market price of rare earth materials for PM is not stable and the figure seems too much values. As for generator losses at copper armature windings and back iron, the value was over 400 kW. If other losses; mechanical, rotor iron losses and so on are added, the efficiency would be decrease around 90 %. This value could be quite lower than that of other conventional multi MW class wind turbine generators.

Secondly, the characteristics of transient state analysis in sudden three-phase short circuit problem have been analyzed. In this analysis, short circuit torque, current and voltage as well as magnetic flux distributions are discussed. The calculation results showed that the transient torque and current values were around 150 % higher than that of rated values.

It was concluded that PMSG structure had many technical challenges for 10 MW class wind turbine generators.

Chapter 4 shows the electromagnetic design of salient pole type superconducting generators (S-SCG) with 2D FEM. Two kinds of electromagnetic characteristics such as steady and transient state are analyzed like the PMSG. For this investigation, two generators with different pole numbers and outer diameters were designed for these two investigations.

The steady state analysis results showed that the diameter of the two designed S-SCG were 5.45 m with 36-pole (S-SCG-A) and 8.2 m with 60-pole (S-SCG-B) respectively. The results explained that S-SCG structure can reduce over 30 percent of generator diameter comparing to PMSG. And the S-SCG-A and S-SCG-B required 62.0 km and 36.9 km respectively. These results showed that S-SCG was suitable to design multiple and larger diameter structure. Focusing on the generator losses, the individual losses were 360.5 [kW] and 278.5 [kW] respectively. If other losses were added to them, they would increase and their values were not so low. However, from the view point of generator efficiency, S-SCG-B would be better choice than S-SCG-A and PMSG.

As for transient analysis, the transient torque was almost the same value of PMSG. However, the transient current value was a little higher than that of PMSG because of lower synchronous reactance. These values should be compared the later two structures.

The S-SCG has a potential for 10 MW class wind turbine generators from the view point of generator costs.

Chapter 5 explains the electromagnetic design of non-salient pole type superconducting generators (NS-SCG) with 2D FEM. Two kinds of electromagnetic characteristics such as steady and transient state are analyzed like the previous chapters. Two kinds of NS-SCG having the different diameter and 12-pole are designed.

The first generator characteristics showed that both generators could obtain 10.0 MW and their diameters were 4.7 m and 8.4 m respectively. They were named NS-SCG-A and NS-SCG-B. They required quite much amount of YBCO wires than S-SCG structures. In case of NS-SCG-A, it required 1240 km of YBCO wire. This was because of its air-cored structure. As for generator losses, the NS-SCG-A had almost 3.2 % of total output; 320 kW and the NS-SCG-B had 480 kW. Finally, the total generator efficiency would be around 90 % like PMSG. From the view point of the efficiency of generators, the NS-SCG structures had the low performances comparing with S-SCG structures.

Focusing on the second transient short circuit analysis, both generators had quite high transient values of torque and current. In case of transient torque, it was over 16 times higher than rated torque; 10 MNm. And also, the maximum transient current values were over 20 times higher than that of rated current; 2,500 A(max).

It was concluded that the generator characteristics of NS-SCG structures had many technical challenges.

Chapter 6 focuses on the electromagnetic design of fully superconducting generators (FSCG) which is the original work of this study. Electromagnetic design and characteristics of the FSCG have been investigated with 2D FEM. As shown in the previous chapters, the analyses in steady and transient state are investigated. Two kinds of the FSCG were designed. One is FSCG-A that the maximum magnetic flux density at the armature winding is 1.5 T. And the other is FSCG-B that the maximum magnetic flux density at the armature winding is 2.0 T.

The first analysis in steady state showed that the amount of HTS wire was less than 200 km, the generator diameter was 4.0 m thanks to the lower air gap; 80 mm in case of FSCG-A. On the other hand, FSCG-B required 260 km with 4.0 m. As for generator losses, this structure had no copper loss but AC loss. In this chapter, AC and iron losses are estimated. Especially, the AC loss in this section means "no load" condition and hysteresis loss of MgB2 superconductor part. The AC losses of FSCG-A and FSCG-B were 1.9 and 1.5 kW respectively. If COP of cooling system and thermal insulation were assumed 0.01 and 100 kW, the total AC losses were estimated at 250-290 kW. These values would be almost 3.0 % of total outputs and not low values for total systems. However, if the armature windings were made with twisted structure, it could have a possibility for AC loss reduction. On the other hand, iron losses in the two generators were 3.0 and 5.9 kW respectively. The ratio of these losses in the total output was only 0.03-0.059 %. These values were quite low comparing with other 10 MW class generators thanks to low frequency like 1.0 Hz.

As for the short circuit problems, the transient torque of FSCG-A was 13.6 MNm. This was almost 1.3 times higher than rated torque; 10.0 MNm. And also, the transient current was 5260 A(max) which was about twice higher value than rated current 2500 A(max). On the other hand the transient torque and current of FSCG-B were 25.5 MNm and 11600 kA(max) respectively. Comparing with NS-SCG, both values of two FSCG structures are reduced dramatically thanks to nonlinear resistance using the I-V characteristics of FSCG contributed to the higher short circuit current area. In other word, when the problem occurred, current-limiting control function based on I-V characteristics of MgB2 had worked at armature windings.

It was concluded that FSCG had good characteristics from the view point of the generator protection from some accidents. FSCG can be a good candidate for the 10 MW class wind turbine generator system.

Chapter 7 discusses the comparison of the electromagnetic characteristics of four designed wind turbine generators in terms of generator diameter, weights, HTS wire length, generator loss and so on. Especially, partial load analysis has been investigated to analyze the generator characteristics with lower wind speed. The results showed that the FSCG and the S-SCG are the best candidates for 10 MW class wind turbine generators from the aspect of size and costs.

Chapter 8 explains the conclusion of this research.

本論文は「Electromagnetic design of light weight and high power density superconducting synchronous machines for 10 MW class wind turbine generators(10MW級風力発電機のための軽量・高出力密度超電導同期機の電磁設計)」と題し、洋上ウィンドファーム等への適用を目指した10MW級の大型低速大トルク風力発電機として、超電導技術の適用可能性を研究したものである。界磁のみを超電導化した2方式、界磁と電機子を共に超電導化した全超電導方式、そして比較対象として永久磁石機を、有限要素法解析を用いて設計し、実用的な超電導発電機設計について検討している。特に全超電導発電機については、MgB2線材を電機子巻線に適用するという新しい提案を行っている。論文は8章から構成される。

第1章は「Introduction」であり、風力発電の現状と大型化の動向、発電機システム構成について整理した後、超電導風力発電機の研究開発の現状と課題について説明し、その上で本研究の目的と論文の構成について述べている。

第2章は「Design Conditions for 10MW class Wind Turbine Generators」と題し、まず本研究で設計、解析する超電導風力発電機の基本諸元と基本構成について記述し、続いて発電機として永久磁石形、突極形超電導、非突極形超電導、全超電導の4種類を具体的に検討すること、さらにそれらの基本構造と界磁コイル、電機子コイル等について説明し、設計および解析の条件を明らかにしている。

第3章は「Electromagnetic Design of Permanent Magnet type Synchronous Generators」と題し、第4章から第6章で設計結果が述べられている超電導風力発電機の比較対象としての永久磁石形風力発電機の設計を、二次元有限要素法解析により行った結果について記述している。定常運転時の発電機特性と短絡事故時の過渡応答特性などの解析結果を示されている。

第4章は「Electromagnetic Design of Salient Pole type Superconducting Generators」と題し、鉄心をできるだけ有効に使った突極形構造の界磁超電導風力発電機の設計結果について記述している。鉄心を比較的多量に使用した多極構造であるため、他の超電導発電機よりも重くなるが、永久磁石形と比較して直径は大幅に小さく、また、使用する超電導線材量も100km以下とすることが可能で、検討した超電導発電機の中で最短であるという結果が得られている。

第5章は「Electromagnetic Design of Non-Salient Pole type Superconducting Generators」と題し、非突極構造でバックアイアンを有する界磁超電導風力発電機の設計結果について記述している。特に高磁界設計を採用し、コンパクト・軽量化を重視した発電機になっている。特に直径を5m以下にした設計では,必要な超電導線材長が1000kmを超えていて、超電導線材のコストを考慮すると、実用性という点で課題のある設計となっている。

第6章は「Electromagnetic Design of Fully Superconducting Generators」と題し、界磁コイルだけでなく、三相電機子巻線も超電導化した全超電導発電機の設計結果について記述している。界磁コイルには他の超電導発電機設計と同様に高温超電導線材を使用する前提であるが、電機子コイルには交流損失を考慮して多心の丸線構造を有するMgB2超電導線材を使用することを提案している。電機子コイル電流が1Hz程度と非常に低い周波数であるため、交流損失は許容可能な程度に収まり、コンパクトで軽量な発電機設計が実現できている。

第7章は「Comparisons of 10 MW class Wind Turbine Generators with Different Structures」と題し、第3章から第6章で述べた発電機設計結果を総合的に比較、考察している。発電機の軽量化やコンパクト化への要求をそれほど重視しない場合は突極形の界磁超電導発電機が実用性の高い設計であり、発電機の軽量化やコンパクト化を実現しつつ、高温超電導線材量も抑えることのできる設計として全超電導発電機の選択があるという結果となっている。

第8章は「Conclusions」であり、本研究の成果を総括している。

以上これを要するに、本論文は、洋上ウィンドファームなどへの適用を目指し、従来の風力発電機の容量を超える10MW級大型風力発電機として、超電導同期機の電磁設計を行い、発電機基本構造と発電機特性との関係を明らかにし、特に新しい全超電導風力発電機を提案して、軽量・高出力密度の特長を有する超電導風力発電機の実現可能性を定量的に示したものであり、電気工学、特に超電導工学に貢献するところが少なくない。

よって本論文は博士(工学)の学位請求論文として合格と認められる。