The medium-frequency power transformer (MFPT) is one of the key elements of an isolated bi-directional dc-dc converters in applications such as future all-dc offshore wind farms, traction, and solid state transformers. This paper describes a design methodology taking into account loss calculation, isolation requirements, and thermal management. Incorporating this design methodology, an optimization process with a wide range of parameter variations is applied on a design example to find the highest power density while the efficiency, isolation, thermal, and leakage inductance requirements are all met.

High power isolated DC-DC converters are likely to provide solutions for many technical challenges associated with power density, efficiency and reliability in potential applications such as offshore wind farms, inter-connection of DC grids, MVDC in data centers and in future solid state transformer applications. The high power medium frequency transformer (HPMFT) is one of the key elements of such a converter to realize the voltage adaption, isolation requirements, as well as high power density. This paper describes a design and optimization methodology taking into account the loss calculation, isolation requirements and thermal management. Incorporating this design methodology, an optimization process with a wide range of parameter variations is applied on a 50 kW, 1 / 3 kV, 5 kHz transformer to find the highest power density while the efficiency, isolation, thermal and leakage inductance requirements are all met. The optimized transformers are then manufactured and will be presented in this paper.

The high power medium frequency transformer (HPMFT) is one of the key elements of an isolated, bi-directional DC-DC converters in applications such as future all-DC offshore wind farms, traction and solid state transformers. This paper describes a design methodology taking into account the loss calculation, isolation requirements and thermal management. Incorporating this design methodology, an optimization process with a wide range of parameter variations is applied on a design example to find the highest power density while the efficiency, isolation, thermal and leakage inductance requirements are all met.

Offshore wind farms pose the challenge of transmission of produced energy in an economic and effective way over long distances, justifying investments in High Voltage Direct Current (HVDC) systems. A solution to surmount the need for an expensive and large power transformer station is to connect each wind turbine to a DC/DC converter in which a Medium Frequency Transformer (MFT) is utilized to reduce the size of the station. Further, the DC outputs of the converters of the wind turbines can be connected in series to create a high DC voltage. In this solution, the MFTs are required to be insulated against a high DC voltage to ground in a limited compartment. This contribution presents a prototype of a 50 kW oil-paper insulated MFT operating at biased 422/4500 V/V AC 5 kHz and 125 kV DC test voltage, which has been manufactured and subjected to verification HV tests. The dielectric design of the prototype MFT based on finite element solution of non-linear Maxwell-Wagner equations is introduced and the acceptance criteria and practical design considerations are discussed.

To automate the procedure of localizing turn-to-turn faults in transformers windings, a method is proposed by employing of Artificial Neural Networks (ANN). For this purpose, a specially made distribution transformer winding is used as a test object to approve the capability of proposed method. This winding is appropriately designed to perform short circuit faults between any two desired adjacent turns. Then the frequency response of winding in both healthy and faulty conditions is measured using the Low Voltage Impulse (LVI) test. Extracted features from frequency responses are used to train and test the proposed ANN. The results show that this method is able to determine the location of turn-to-turn fault in winding.

Power transformers are considered as one of the essential elements in electrical networks. Any failure in these equipments directly reduces network reliability and increases maintenance costs. Therefore, quick and precise diagnosis of faults is very important. One of the main faults in this regards is turn-to-turn fault that its localization is very useful and vital. This paper has been presented a novel method for diagnosing and localizing of above mentioned faults in power transformers using both transfer function method and neural network technique. In this contribution some different turn-to-turn faults have been simulated on a laboratory HV winding of power transformer and correspond their transfer functions have been measured. Then, using an ANN with Back-Propagation learning algorithm, a method has been proposed to identify the local of occurred fault. Results show that the ANN could well diagnose the local of fault if it is trained using the presented patterns.