by Dr. Dan Martin, Australasian Transformer Innovation Centre, the University of Queensland
Network companies face growing challenges managing their aging power transformers whilst delivering high network reliability at lower costs, with increasing customer expectations.
There is a rapid shift towards large-scale variable renewable generation and disruptive new technologies such as peer-to-peer retail trading, batteries and high-speed electric vehicle charging. Power transformers must operate reliably for several decades and withstand these rapid changes that significantly impact their loading.
Risks to aging power transformers
Australia’s fleet of power transformers have already, on average, exceeded 70 per cent of their original design life of 30 to 50 years, depending on operating conditions and desired reliability. In the current economically-constrained environment, Australia’s regulators are demanding higher transformer utilisation and even longer lives with higher network reliability.
Distribution and transmission companies must respond to the regulator’s and customer’s demand by adopting innovative transformer lifecycle management practices and new condition monitoring technologies to assess and manage transformer failure risks and extend transformer life.
This is a challenging task, requiring experienced staff equipped with innovative tools, comprehensive data and sound engineering judgement.
Low cost approaches are needed to gather essential data from aging transformers to including up-to-date measurements of oil impurities, partial discharges etc. Whilst online monitoring may be a theoretical option, it can be expensive and difficult to economically justify.
2015 age distribution of Australian power transformers
A compounding factor is the difficulty in justifying the return on any investment in conditioning systems near the end of a transformer’s life when its residual value is quite low.
The network is changing. Transmission and distribution networks globally were designed to transfer electrical power from large fossil-fuelled generators to load centres. In Australia, these generators were located near remote low cost, coal mines, linked by a high-capacity transmission system to distant distribution networks.
In contrast, most new solar farms are being developed at the western and northern extremities of the grid where solar insolation is higher and land is cheaper. The existing grid in such areas has low capacity because of the sparse population.
The network company is faced with the challenge of significant increases in network loading and its variability combined with minimal support for network reinforcements.
Australia’s regulatory framework requires shared network upgrade costs to be met by electricity users, other than dedicated radial connection assets which are funded by the new generator. Regardless, the network company must transparently justify investment in new assets.
Changing customer behaviour
Two new disruptive technologies are electric vehicles and transactive energy. Electric vehicles are on the cusp of becoming mainstream, with many of the global auto-manufacturers releasing their plans for mass roll-out in the very near future. High-current EV chargers will place a strain on the supporting distribution network, especially distribution transformers, reducing their residual life.
Transactive retail energy trading, commonly called peer-to-peer trading, will see households making greater use of the distribution networks to trade electricity. Low and medium voltage distribution networks have almost no monitoring because historically this was not required, there was only a traditional one-directional flow of electricity towards every house, and the large distribution networks were made cheap to simplify designs.
However, the introduction of rooftop PV and the new disruptive technologies will lead to higher bi-directional power flows on these networks overloading distribution transformers and causing transient overvoltages on LV feeders.
Experimental solar farm operated by the University in Queensland
Future for power transformer operation and transitioning to a data-rich future
The increasing requirements to monitor the loading and condition of power transformers as well as increasing transformer utilisation and effective life will require new systems to gather, analyse and interpret asset data to improve asset oversight, and reduce the likelihood of unforeseen asset failures.
Online monitoring has been available for some time, however due to initial investment costs it has only been used on high value, high risk transformers. With falling costs and improving reliability many new monitoring technologies are being considered which can provide data on the condition of the transformer.
However, this ICT infrastructure will require management by the utility, and then the usual considerations apply on keeping software updated and managing hardware obsolescence.
One strategy to reduce management costs is to keep the monitoring system as simple as possible, and use centralised computing for processing. Cloud-based software is becoming increasingly attractive to process the incoming data on transformer usage and condition to provide useful information on the assets, and there is now more focus on cyber security.
Research into transformer condition monitoring at the Transformer Innovation Centre has included the development of thermal models to understand the effects of short-term overloading on reliability and residual life. The more detailed the mode, the more data is required and the greater the challenge to align internal business processes and decision making with the analytics and data.
Leaving aside the initial setup cost of data analytics, a 2016 McKinsey & Company report identified a 36 per cent saving in utility maintenance costs. Given that the Australian transformer fleet costs $200 million p.a. to maintain, the potential savings are significant. It is always best to focus on areas that deliver the ‘biggest bang for the buck’, which requires expert knowledge to identify.
The Australasian Transformer Innovation Centre (TIC) was set up with the objective of supporting industry to close the gaps between aging assets, new technologies and future needs of the network and stakeholders.
Established in 2017 at the University of Queensland by a consortium of three universities and 19 industry members, TIC researches the transformer technologies identified by its members, and delivers industry-focused training courses on key aspects of transformer lifecycle management and practical condition monitoring.
The following four TIC research projects are already underway:
- Resilience of networks using vegetable oil-filled transformers. Previous work suggested that power transformers filled with vegetable oil can be operated at higher temperatures than normal mineral oil. This work focuses on periods of extreme loading, such as during heatwave events, and how to maintain electricity supply to consumers
- How asset management of transformers should change with the incoming wave of large-scale solar generation often located at the extremity of the network. Higher levels of harmonics may decrease the residual life of the existing transformers and asset owners must mitigate potential problems
- Optimising techniques to measure partial discharge within power transformers to predict impending failure
- Potential power quality problems caused by high-current power electronic devices that can affect power transformers
Three continuing professional development courses have already been developed since September 2017. The first one was a two-day introductory course on power transformer lifecycle management. The second was an advanced course on transformer bushing design, maintenance and operation. The third, which will be delivered on 27-28th June 2018, is an advanced course on power transformer tap changers: design, maintenance and retrofit. These courses are largely delivered by industry experts in addition to TIC university staff and all attendees rate these TIC courses as excellent.
For more information visit www.itee.uq.edu.au/tic.
