Environment & Engineering

Electrical Fault Identification

AdobeStock 236082832

The electricity distribution network across New Zealand is extensive with 98,549km of overhead distribution powerlines, yet it is around 70 years old and would cost a significant amount to update. With increasing demand on the infrastructure caused by new technologies such as electric vehicles and renewable energy generation, the aged network is prone to many different faults including unplanned outages.

Anwarul Sifat and his team at Robinson Research Institute are developing a non-contact magnetic sensor, which can detect power faults across electricity networks. Distribution companies across the country do not currently have visibility over line activity and unplanned outages or damage can cost significant amounts to repair. This device will allow companies to capture data through the sensor and analyse this using machine learning techniques to classify the fault type based on patterns.

Features and Benefits

Non-intrusive, compact sensor

The sensor can be used to estimate the overhead line currents or load current rating in a non-contact way. Therefore, it can be easily added to the distribution network, rather than having to rehaul the whole network, which would be incredibly costly. The sensors are compact, cheap, accurate and portable.

Overhead line management

The machine learning algorithm assessing data from the sensor can classify several different types of faults, including faults that have not yet reached a critical stage. This enables distribution companies to make decisions about how best manage their overhead lines.

Increase safety

A broken conductor can be incredibly dangerous as electricity continues to run through it. The device can be used to predict areas across the distribution network that are prone to high impedance faults, reducing the amount of damage.

Next Steps

Sifat is one of the newest recruits on the KiwiNet Emerging Innovator Programme and is connecting with industry partners to develop his technology to be used across the distribution networks.

Currently, he is involved in customer discovery and market validation research. In addition, he is negotiating with leading local distribution power supply companies to install a prototype of his sensing systems to collect real-world fault data.

To find out more, please contact the Commercialisation Manager, below.

Hamish Findlay 2 v2
Hamish Findlay

General Manager Commercialisation

View Profile
Sam Wojcik
Sam Wojcik

Commercialisation Manager

View Profile