This thesis describes four novel superconducting machine concepts, in the pursuit of finding a suitable design for large offshore wind turbines. The designs should be reliable, modular and light-weight. The main novelty of the topologies reside in using a single loop shaped stationary superconducting field winding, which eliminates the rotating transfer couplers and electric brushes or brushless exciters. Furthermore, the electromagnetic forces in the superconducting wire are also eliminated, which simplifies the design and manufacturing of the cryostat and the support structure. Among the four topologies presented, the claw pole type machine is the most promising one. The rotor of the machine composes of claw-poles made from laminated electrical sheets, the superconducting field winding and the armature winding are stationary. The machine is analysed using 3D FEA simulations and a small linear machine prototype is manufactured to verify the simulations. For large scale applications, a double-sided claw pole machine is proposed, which has balanced magnetic attraction forces in the rotor. The machine has a modular cryostat structure, which increases the availability of the machine. Thus, even if a fault occurs in the cryocoolers or in the armature coils, the rest of the machine can operate at partial load until the maintenance is performed. Moreover, it is much easier to replace the faulty parts, as full disassemble of the machine is not required, and a small on-site crane can be used. As a result, it offers operational advantages over the existing superconducting topologies. A 10 MW, 10 rpm generator design is presented, which has a diameter of 6.6 m and an axial length of 1.4 m. The total active mass of the generator is 58 tonnes, and the structural mass is 126 tonnes, which gives a total mass of 184 tonnes. There are four independent cryostats and two independent armature windings in the machine to improve modularity. The biggest advantage of the design is the significantly less superconducting wire usage compared to any other designs; 10 MW machine just needs 15 km of MgB2 wire at 30 K. Thus, it is believed that the proposed topology is a very cost effective and suitable candidate for a successful entry to the wind turbine market.