GE Digital Energy have a range of products which provide diagnoses for power utility and grid assets. Critical assets include high-voltage (HV) circuit breakers and power transformers. Sudden failure of utility assets can have catastrophic consequences for electricity security and may lead to grid collapse.Circuit breakers are maintained at regular intervals, however transformers are sealed units and diagnostics are 'black-box' based using external electrical/chemical measurements. Although transformer lifespans typically 40 years, this can be extended a further 10 - 20 years if properly maintained and strict condition monitoring applied. Diagnostics can provide essential information about transformer health and provide early warning of potential problems, but while there is much published research for specific diagnostics, there is no accepted measure or benchmark to reliably predict power transformer lifespan. Since transformers vary widely in design, operation and connection, the development of a standard approach for diagnostic prediction to assess transformer fleetlifespanwould offer significant benefits for manufacturers, power utility maintenance and equipment suppliers.
The overarching aim of this project is the construction of an approach for power utilities to use transformer fleet data to predict lifespan and best direct maintenance investment. The objectives of this project are to: 1) examine variation in transformer construction and integration and devise condition monitoring parameters using data collation; 2) investigate models for accurate prediction/diagnostics using external measurements, e.g. dissolved gas analysis (DGA), partial discharge and load profiling; 3) devise and validate a general model and tune parameters using historical fleet data; and 4) develop algorithmic tools for pre-emptive diagnostics/prediction fortransformer fleet management applications and deployment.
Project research aligns well with the School of EEECS academic standards and will help promote the local, national and international reputation of the School and further the strategic objectives of the Energy, Power and Intelligent Control (EPIC) cluster. The project will be undertaken by a PhD student and contribute to the School objective of increasing postgraduate research recruitment and fulfil the Cluster target of developing sustainable research, building expertise and impact. The School/Cluster will significantly benefit from engagement with GE Digital Energy through knowledge exchange. Project objectives will be strategically enhanced by strong Partner collaboration through rapid-start access to industrial expertise in transformer diagnostics and engagement with related projects.
The project will generate rigorous, original and high quality research output for scholarly dissemination and develop the societal impact of QUB research output by providing a potential case study for the 2020 Research Excellence Framework (REF) submission. The project will provide GE Digital Energy with access to local QUB resources, academic expertise and render a proof-of-concept algorithm for further development and integration. Outcomes with economic potential will clearly merit commercial investment by the project Partner.
The proposed area of investigation in transformer diagnostics also aligns well with Cluster research and outcomes and will support the academic reputation of the School/QUB and provide international recognition through publication. The project will provide the Partner with timely outcomes for products which benefit power utility operators responsible for asset lifespan management and maintenance.