Curtin University is seeking commercial partners to collaborate towards the development of a product based on this unique technology.
Email: commercialisations@curtin.edu.au
Summary of technology
Power transformers are used in the electrical grid to step up and step down voltages for transmission between power stations and consumers of electricity, namely residences and businesses. As major parts of the distribution network, they are susceptible to failure due to external damage or general aging of parts.
Transformer failure interrupts power delivery to the end user and can lead to considerable replacement costs. Catastrophic failure may result in explosion and fire, potentially incapacitating the network and threatening property and lives.
Curtin University researchers have developed an algorithm to continually monitor power transformers for internal mechanical and electrical faults that may result in loss of performance or dangerous breakdown. The method calculates the relationship between the input current and the difference between input and output voltage, then compares the resulting ellipsoid graph with what is expected of a healthy transformer.
We have developed the diagnostic side of this test to detect minor changes in the ellipse, which represent a range of different faults that may have occurred in the transformer. Faults include:
- turn-to-turn short circuit
- disk deformation, displacement or buckling
- disk to ground and electrical leakage faults
- axial displacement.
Methods that are currently used for transformer fault detection include Frequency Response Analysis (FRA) and Dissolved Gas Analysis (DGA). FRA suffers from a requirement to disconnect the transformer before testing, which interrupts local power supplies. DGA is used in an online monitoring mode, but is limited to detecting only those faults associated with partial electrical discharge.