Case Study

Floods of alarming insights

Advanced Analytics | Custom Analytics + Software Solutions | Data Visualisation + Reporting Solutions

A major Energy Networks company in Australia was concerned about the rapidly growing number of alarms on their distribution network.

The Challenge

Within the control room, controllers were swarmed with alarms every second, needlessly elevating stress and lowering morale. While the notification and monitoring of alarms is crucial to a safely operating network, a majority of the alarms were not actual problems, and were consuming valuable time and resources. Additionally, a specific type of device was introduced, dedicated to the real-time detection of faults that could ultimately lead to bushfires. These devices are tested every 10 minutes, which causes a flood of alarms across the network and further stresses a busy control room.

The company wanted to address the root causes of alarm growth and ultimately reduce the number of alarms over time. The dashboard Harmonic designed and built will provide insights to support key decisions on filtering, grouping, and removing alarms within their distribution management system.

The Solution

Harmonic developed a PowerBI dashboard using data from the company’s distribution management system, GE PowerOn Fusion. The dashboard’s role was to provide crucial statistics (such as alarm rate), visualised trends, and classify alarms into respective categories (including fleeting and chattering noisy alarms). To investigate the pain points within the network, we also included options to sort and filter the data spatially, temporally, and categorically. Additionally, functionality was also included to monitor the controller’s use of the dashboard, including tracking the responsiveness of controllers to actioning alarms, asset configuration and alarm definition configuration.

In addition to the core visualisation dashboard, the client sought to identify alarms which were linked to noisy alarms, so they could be investigated for causal links. To address this request, machine learning techniques were used to mine the data for association rules. These rules were used to categorise alarms which were not obviously related with the bushfire fault detection device, but nonetheless tightly linked, so our client could diagnose their root cause.

The Result

  • Improving categorisation and definitions of alarms led to a reduction in noisy alarms.
  • Improved responsiveness to alarms reduced customer impact and costs.
  • Fewer, and more meaningful alarms, improved controller morale.

The modern Control Room must focus on Situational Awareness to support the encroaching complexity of daily operations that are being impacted by new and irregularly patterned events on the team responsible for critical decision making.

Analytical focus on Alarms has added an extra dimension to the modern problems to be solved, and provided rich insights to support next stage investments, plus the confidence in the impact they will have when delivered.

— Energy Platforms Manager

Get in touch

Please provide your first name.
Please provide your last name.
Please provide the company you are currently working at.
Please provide your current position.
Please provide your work email address.
Please provide a valid phone number.
Please provide a message.
captcha
Invalid Captcha. Please try again.