Australasian Science Magazine

The sensors can be mounted on power poles tens of kilometres apart and used to locate a fault point and thus detect the incipient signs of a blackout and help to prevent it.

By Alexe Bojovschi & Khoi Loon Wong

Electric fish have inspired a new technology that can wirelessly detect where faults in the energy grid are likely to occur before they lead to blackouts.

Although many factors have contributed to the recent increase in the number of power outages, two predominate – ageing infrastructure and increasing environmental pollutants. Both will remain a problem without the development of an efficient distributed technology that can detect incipient signs of power failure.

The majority of the power networks were built at least 50 years ago, and they are ageing and deteriorating just at the very time when they are being overloaded with new appliances. This increases the probability of the development of sparks. All it takes is a salt deposit or a build-up of lichen to provide a conductive path across an insulator and increase the likelihood of electrical discharges. Small electric sparks can evolve over time into larger electrical discharges. This dynamic process is now widespread in power networks and leads to outages.

In 2012, blackouts left more than 620 million people in India without power for a couple of days and in the USA cost the economy more than US$120 billion. Electric sparking has been blamed for major bushfires in Australia.

The psychological significance of blackouts is immense. Just imagine being alone in a lift or even at home when the lights go out. You might feel irritated, angry, terrified or lost, but unless you have your own generator you will also be ... powerless. More serious, of course, are the difficulties for emergency facilities, such as hospitals, where people’s lives depend on a continuous power supply. The wireless sensing technology that we have developed attempts to prevent these sorts of events from happening.

At present, power grids are being challenged by new requirements such as those related to the integration of renewable sources like solar panels. The development of a technology that allows two-way electricity and information flow between power utilities and consumers is required. Such technology allows integration and management of renewable energy generation and consumption while at the same time providing increased infrastructure security. It also assists appropriate management of high-demand across the grid.

We carried out fundamental research that has converged to ensure a safer power network. The work, which has been patented and published in a number of engineering and physics journals, is based on electric fish.

Electric fish have poor eyesight and instead rely on electrical organs to transmit and sense electromagnetic signals. This allows them to detect prey and to navigate through murky waters. The signals that electric fish generate and detect are similar to those emitted by power network components.

The frequency and intensity of the electric field around the body of an electric fish depends on its size and on the characteristics of the surrounding environment, such as changes in the chemical composition of the water. For example, the concentration of CO2, and hence the acidity of the water, are major factors that trigger different frequency responses by the fish, which can generate a range of signals – from short electrical pulses with narrow frequency bands to broadband signals.

We have translated the way an electric fish responds to its environment and senses the characteristics of objects around them to the power network. We studied the patterns of electro­magnetic radiation emitted by faults and electrical discharges in the power network and determined a typical signal, how it was transmitted and how it interacted with the components of the grid. This knowledge constitutes the cornerstone of our wireless technology for detecting the early signs of power failure.

We have developed sensors to detect the electromagnetic signature of electric sparking. The sensors can be mounted on power poles tens of kilometres apart and used to locate a fault point by translating the time of arrival of the fault signature into a measure of distance. The difference in the time of arrival of a signal at different sensors is used to derive the distance, and hence the location, of electric sparking discharge. In this way the technology is able to detect the incipient signs of a blackout and thus help to prevent it.

Our technology has been deployed in Victoria’s power network, and plays a key role in ensuring infrastructure security. It opens the way to a distributed sensing network for maintaining a smarter grid.

We have established a company to commercialise the system. At present, IND Technology (www.ind-technology.com.au) is offering the technology as an early-fault-detection service to electricity companies in Victoria online 24 hours per day. The system provides them with a dynamic picture of the health of their power networks.

Because the degree of discharge is directly related to the remaining life of high voltage components, we can predict how long insulators, generators, transformers and transmission and distribution lines are likely to last. This helps with maintenance and efficient management of the network.

We have found that the conditions that lead to discharge events and blackouts are diverse. The main ones are contamination build-up, wetting of the surface, leakage current, resistance heating, spot discharges and deterioration of insulating materials. These can all be related to the level of pollution. So, one of the questions arising from this is whether the insulators in the power network should be optimised for the level of pollutants likely to be encountered in heavily industrialised cities.

The fundamental research incorporated in this technology involved the use of advance supercomputing methods and experiments. This allowed us to correlate the level of discharge activity and electromagnetic radiation with the level of chemical pollutants, rain density, wind speed, atmospheric pressure, leakage current, thermal heating and the dimensions of the discharge events. Besides the metallic components of the wires, four common dielectric materials that are used by the power industry – polymers, epoxy resins, ceramics and wood – were considered. We discovered that the radiation pattern emitted from sparks on a high voltage insulator is correlated with the intensity of the source and the condition and geometry of the dielectric material.

The knowledge from these studies provides a unique capability for predicting the characteristics of power network components with unsurpassed accuracy. Defects develop over their lifetime and can be triggered or accelerated by diverse factors. Monitoring of electrical sparking provides an accurate indication of the condition of insulation.

To guarantee a high level of reliability, continuous monitoring of the network is required. This will be paramount in our electric future.

Alexe Bojovschi and Khoi Loon Wong conducted this research at RMIT University’s School of Electrical and Computer Engineering. Khoi Loon Wong is co-founder of IND Technology.