AWARDS FINALIST: Northpower – creating thriving communities

4 Jul 2024

Northpower drew heavily on new technology over the past year to address legacy safety issues on its network while also preparing for the rapid upgrade and expansion of its operation during the coming decade.

The company has just delivered a record network spend of $40 million, a 52 per cent increase in capital delivery on the previous year.

Faced with a growing demand for capital investment, community owned Northpower also adopted a new investment governance framework in the past year to challenge and justify the value to customers of every dollar spent.

The company says the new regime has tightened and streamlined its control of spending and minimised resources going into projects that present well in the concept stage but don’t measure up under more detailed examination.

The company also partnered with Australian modelling company Modla to improve its asset management planning to enable specific intervention for individual assets.

Analysing data

It has also taken on the Gridsight tool being used by other distributors to analyse smart-meter data to improve network utilisation and capacity. It has now piloted its own project and is about to contract for the full analytics piece of the project.

Customers will also benefit from an improved distribution management system achieved by integrating fault ticketing in the firm’s customer relationship management system with its outage management system. Real-time information from customers informs its field force, while the OMS link allows the CRM to provide automated notifications of planned and unplanned outages, and real-time information for the website and call centre staff.

Northpower also deployed artificial intelligence and machine learning to improve safety after the death of a cow highlighted the risk of live stay wires on its network.

The condition of stay wires on more than 60,000 poles needed assessing but posed the challenge of interrogating 70,000 scanned historical inspection reports and 800,000 inspection photographs of varying quality.

Northpower developed in-house a machine-learning approach which combined optical character recognition (OCR) and computer vision (CV) techniques to extract the required information.

A team of volunteers then annotated 10,000 photographs, which were used to train the machine-learning model to recognise and localise these features within the photographs.

Northpower says it was then able to categorise poles into high, medium, and low priority groups. Thirty-one poles were identified without insulators in stay-wires, and an additional 110 poles required further field investigation.

The trained model also identified potheads: pole-mounted legacy equipment prone to explosive failure. The CV model found eight more potheads than were already known.

Northpower says the system has significantly improved the visibility of its historic data records, improved safety and provided a more complete picture of its asset base.

Decommissioned lines

The firm also took a novel approach after another livestock death that highlighted the potential risk from inadequately decommissioned service lines.

It employed a 12-strong student army to assess the condition of about 4500 decommissioned lines over four months and then verify the findings with field inspections over the previous summer.

The project, with technical support from the firm’s engineering team and mentoring by senior staff, identified more than 600 customer service lines to be cut away from the network and a handful that requiring immediate attention.

This project was an opportunity to inspire a group of soon-to-be engineers and provide mentoring and exposure to the electricity sector. The application developed by our digital team can now be shared with others facing similar issues, the company says.

A refreshed “stay safe together” public safety campaign raised public awareness of the stay wire issues alongside the AI work which sped up the response.

Meaningful customer engagement and education continues to be a high priority for the firm, with various customer outreach programmes and supporting those in greatest need through a home energy education programme.

The Energy Distributor of the Year category is sponsored by Axos Systems.