Engineers use data to manage grid transformers, boosting reliability to homes, farms

AMES, Iowa – Pay attention the next time you drive near your home, farm or business. You’ll notice small, green utility boxes all over the place. They’re distribution transformers. If they’re not working properly, electricity won’t flow to your lights and appliances.

 

Those boxes take kilovolts of electricity (that’s high voltage, measured in 1,000s of volts) from transmission lines and step it down to the safer, practical 120 or 240 volts that power our daily lives.

 

“Utilities have plenty of them,” said Zhaoyu Wang, an Iowa State University professor of electrical and computer engineering. “Most of them only supply two to 10 customers.”

 

The city of Ames, for example, with a population of about 66,000, has about 5,500 distribution transformers on its grid serving about 29,000 customers, according to the city’s Electric Department.

 

These are not smart devices. There are no sensors attached to let utilities know if there’s any kind  of problem. Utilities have been in the habit of keeping a large inventory of the boxes that had cost $1,000 to $2,000 apiece.

 

But that’s no longer a good option. Costs of the boxes have tripled. Boxes are on long back orders. And the boxes are getting overloaded and overheated as we all depend on more and more electricity to run vehicles, heat pumps, tools and devices.

 

“Every time the temperature of distribution transformers goes up, the lifetime of the boxes decreases,” Wang said.

 

Wang has an idea to fix that burnout problem, one that would help utilities move from a “passive ‘broken-fix’ cycle” of managing distribution transformers to a “proactive ‘monitoring-prediction-maintenance’ cycle,” according to summary of his research. That idea could minimize service disruptions while advancing distribution reliability and resilience, while also lowering grid capital and operating costs.

 

A more reliable grid for you

The U.S. Department of Energy’s Office of Electricity recently announced a $7.5 million initiative to support eight research projects that use data and sensor technologies to boost grid reliability and resilience.

 

 

The initiative awarded Wang and his collaborators a three-year, $1 million grant to support their work with distribution transformers. Other co-leaders of the project are Anne Kimber, the director of Iowa State’s Electric Power Research Center; and Bai Cui, an assistant professor of electrical and computer engineering.

 

Project partners also include the National Renewable Energy Laboratory in Colorado; the Linn County (Iowa) Rural Electric Cooperative; Cedar Falls (Iowa) Utilities; the AES Corp., a global energy company; and SparkMeter Inc., a global provider of grid management solutions.

 

The researchers and partners will work together to develop a data-driven system that monitors the loads on distribution transformers and gives utilities real-time awareness of which boxes should be checked and potentially upgraded or replaced.

 

So how can the researchers come up with a data-driven system when distribution transformers have never produced any data?

 

Many of the homes and businesses connected to distribution boxes have smart meters, Wang said. Every 15 minutes, those meters report energy consumption and voltage back to utilities.

 

“If I know this, I can get load information of distribution transformers every 15 minutes and can then estimate operating temperatures,” Wang said. “That allows us to estimate degradation and remaining useful life. And then we can provide a ranking of distribution transformers to evaluate.”

 

Machine-learning technology will analyze the data to indirectly determine the health of distribution transformers on the grid.

 

Wang said the system will rely on SparkMeter’s artificial intelligence-driven Praxis platform and be integrated into its commercially available GridScan product, making data monitoring and analytics readily available to utilities.

 

All of that, Wang said, can help the energy department meet its goal of using data to help utilities quickly identify and solve problems. And for utility customers, that means a more reliable power supply for everything they’re plugging into the grid.

 

– 30 –

---

---