Duke Energy: Driving more informed power transformer maintenance decisions through online multi-gas DGA monitoring

How one of America’s largest energy holding companies, Duke Energy, is leveraging Vaisala’s online multi-gas DGA solution for power transformer monitoring to improve asset management and drastically reduce unintended labor costs and unplanned outages.

The challenge:
How to effectively and efficiently monitor transformers that require frequent sampling and deeper analysis of potential faults

Duke manages hundreds of substations that include thousands of transformers, some of which have a lower voltage rating and come at a low enough price point that they do not warrant permanent online DGA monitoring. This includes many medium and even some higher voltage transformers. For these transformers, the current monitoring process involves annual or bi-annual sampling. The frequency is adjusted based on those results and in some cases can involve a weekly sampling schedule, dramatically increasing O&M costs. 

Duke has a well-developed monitoring and sampling protocol in place. But while the system was recognizing that some of the smaller transformers in its portfolio were potentially experiencing issues, the only solution was to have field personnel take more frequent oil samples to try to detect gassing patterns and potential faults. This was an expensive and time-consuming process for the maintenance crews. The cost of collecting and analyzing an oil sample from a transformer is significant, with decision-makers having to wait up to 24 hours to get the data back. 

Duke needed to find a safer and more cost-efficient solution to address the viability of its transformers. 

This case reviews two specific transformers that were monitored with Vaisala’s mobile OPT100 online DGA. The first being a single phase where the other phases had already failed. The second being a unit that had concerning gas results, and was critical to the local network with no spare or way to offload should a failure occur. 

The solution:
Mobile online DGA monitoring for real-time analysis

Seeking a solution that enables maintenance crews to analyze oil on-site instead of sending it to a lab for DGA information, Duke Energy piloted Vaisala’s mobile OPT100 monitoring solution at a substation on Eagle Island outside of Wilmington, NC, for a four-month project beginning in December 2020. 

Leveraging Vaisala’s Optimus™ technology, this solution enabled a maintenance-free option for measuring fault gases in transformers. The core NDIR system is based on a proven technology Vaisala has used for decades and enables the OPT100 to deliver reliable long-term measurements. In addition to no maintenance or consumables required, the OPT100's robust design and ease of installation (under two hours), make it uniquely positioned to be offered on a transportable platform that is available in the U.S. market. 

The Optimus provides real-time fault gas monitoring, including, methane (CH4), ethane (C2H6), ethylene (C2H4), acetylene (C2H2), carbon monoxide (CO), carbon dioxide (CO2) and hydrogen (H2), enabling a deeper analysis of faults using analytical tools like the Duval triangles, key ratios or analytical software that compares DGA to load and other factors in real time. This cannot be done as effectively with offline lab sampling/testing.

After installation at the first site, the data collected proved the transformer was stable with no issues needing to be addressed. Taking advantage of the ease of mobility of the OPT100, the Duke substation maintenance team decided to move the unit to a more concerning transformer. Since the substation in question was located an hour away from the closest maintenance facility, Duke also saw added time and resource savings benefits. They no longer needed to send maintenance personnel out to collect samples, driving out and back and then filling out paperwork before the sample could be analyzed. 

As part of the project, Duke needed to be able to collect and analyze the DGA data on a regular and continuous basis. By gathering sufficient real-time intelligence, they could then effectively interpret any issues. This was particularly useful, as Duke had O&M resource limitations.

Interpreting the gassing patterns and severity was not possible in previous years, as the sampling frequency was too spread out. This made the actual gassing patterns difficult to decipher and the exact event timing challenging to determine. By having the online monitor continuously measuring the dissolved fault gasses and outputting the data every hour, the utility benefited from precise and reliable real-time data for trend analysis and correlation to other factors, from load and temperatures to PD. 

The benefit of real-time visibility to save time and money on operations and maintenance 

The last manual sample collected from the nearly 70-year-old critical transformer at this substation showed significant increases in ethane, acetylene and hydrogen. Two weeks following the installation of the mobile OPT100 at the new substation location, the utility began to see gassing events on the unit. These events continued over the following four weeks. 

With the mobile OPT100 online monitoring system installed and delivering data, the Duke maintenance team was able to: 

• remotely track and trend the gas data quickly and easily.
• compare and integrate online DGA monitor data with traditional lab sample data to provide a cohesive history of transformer DGA tests.
• completely diagnose the transformer at hand, including identifying anomalies and gassing events overlooked by the traditional sampling practices.

“The monitor’s main selling points are that it is mobile, easy to deploy and provides accurate real-time information on which we can make informed decisions,” said Adam Long, lead transformer engineer. “The connectivity and remote means of accessing data prevent our people from [having to] driving out to collect data because the device sends the information for remote analysis.”

With the OPT100 installed and monitoring, Duke soon saw a continuous increase of acetylene – the cumulative, incremental growth – which was verified by manual samples during the trial. Duke was then able to leverage that visibility into the transformer’s acetylene trend to recognize an issue that needed to be addressed in the short term. Luckily, the transformer in question already needed to be de-energized for maintenance activities to be performed and was on the short-term replacement list.

However, the monitoring data was vital information to inform the maintenance team to accelerate the maintenance plan and, more importantly, move its replacement timeline up by two years — thereby avoiding a potential failure and the costs of an unplanned outage. The online monitor was also key in serving as a safety check as Duke had to wait a year before it could take the transformer offline permanently.

Conclusion:

The mobile Online DGA monitor proved its value as an important tool for Duke.  With the ability to monitor and compare gassing trends in real time, decision-makers could prioritize employee safety while significantly saving time and money regarding operations and maintenance.

Specifically in this case, the accurate and reliable data gave maintenance crews the peace of mind that this specific transformer was not a bomb about to explode and was safe to keep operating while in fault-monitoring status until the utility was able to go another transformer installed. 

While offline DGA sampling will identify potential internal faults, only continuous online DGA monitoring tools like the OPT100 can provide real time DGA data that reveals the actual gassing patterns, enabling true fault identification and comparative analysis.

The technical life and current condition of the transformer — particularly if a fault requires immediate maintenance action or replacement — are highly valuable to asset and maintenance managers who need to make timely maintenance action decisions, so having a mobile online DGA monitor is a key tool.

At the end of the day, there is nothing worse than an unplanned outage — especially when that outage could have been avoided by leveraging online monitoring tools to detect serious power transformer faults.

If you have any questions about this case please do not hesitate to contact us.

For a full view of Vaisala's offering for the Power, click here.