How artificial intelligence helps ensure the security and reliability of America’s energy infrastructure

Argonne scientists are harnessing the power of artificial intelligence to transform the maintenance of power grid assets, helping U.S. utilities identify and solve problems before they occur.

America’s energy demand has never been greater and continues to grow – recent filings with the Federal Energy Regulatory Commission show that grid planners expect demand to grow by almost 5% over the next five years. To meet future energy needs, utilities must launch new power plants while maintaining existing infrastructure.

Renewable energy sources such as wind, solar and hydropower will play an increasingly important role. The United States aims to generate 44% of its energy from renewable sources by 2050, more than doubling the power currently produced by these new technologies.

Solar energy is expected to provide 22% of our energy, with another 14% coming from wind. Integrating these new energy sources into the grid will require the installation of hundreds of millions of inverters, each of which will require maintenance.

Meanwhile, parts of the existing network are old and starting to fail. The average age of hydropower facilities in the U.S. is over 70 years. Many are reaching the end of their service life and require extensive inspection and maintenance. Even more troubling is the state of America’s power lines, supply networks and gas pipelines. The American Society of Civil Engineers gave these systems a C- rating for 2021.

Monitoring and maintaining the health of this diverse set of energy assets of varying ages is critical to ensuring the reliability and security of our electrical grid. However, utilities may not know there is a problem with their equipment until something breaks.

To meet this need, scientists at the U.S. Department of Energy’s Argonne National Laboratory are stepping in. Working closely with utilities across the energy sector, from aging hydroelectric plants to massive solar installations, they are changing the way companies approach maintaining the nation’s energy infrastructure and clean energy assets.

Using the latest artificial intelligence (AI) technology, Argonne researchers have developed AI-enabled software that can predict network component failures. The system analyzes the vast amounts of information that utilities collect from sensors installed across the network, creating a predictive model that forecasts consumption over time. Ultimately, the software can recommend when to repair or replace parts before any problems occur.

“Companies want to know what the health of their assets is,” said Feng Qiu, head of Argonne’s Advanced Grid Modeling group, which led the research. “Our predictive models using condition monitoring information can tell them how much useful life is left on their equipment – ​​how many years, months and weeks it has left.”

Shijia Zhao, an energy systems scientist at Argonne who played a key role in the research, explains that their approach goes beyond traditional reactive maintenance strategies. “Instead of waiting for equipment to break down, we use artificial intelligence to proactively identify potential problems and schedule maintenance in time, saving utilities time and money.”

At the heart of this innovative approach is the ability to estimate the condition of infrastructure and assets, predict failure risk and adapt maintenance decisions based on current real-world data. By moving from laboratory models to data collected in the field, Argonne researchers showed how useful this technology can be for energy suppliers.

In one solar inverter project, the team showed that it could potentially reduce overall maintenance costs by 43-56%, unnecessary crew visits by 60-66%, and increase profit by 3-4%.

“Our goal is to equip energy providers with the tools they need to ensure a reliable and resilient grid for years to come,” Qiu said. “With this technology, companies can make informed decisions about when and how to repair or replace equipment, ultimately improving the overall efficiency, safety and reliability of America’s energy infrastructure.”

The benefits of this study go far beyond savings and efficiency gains. By minimizing outages and resolving maintenance issues before they escalate, energy providers can increase grid reliability and resiliency, a key factor in an era of growing energy demand and a changing energy landscape.

The power and scale of AI-powered prediction and optimization models means they can optimize maintenance at the network level. “It’s extremely important to keep the lights on,” Qiu said.

By looking at the entire electrical grid – from power plants to power lines – models can predict failures across the entire network that produces and transports electricity from where it is produced to where it is consumed.

There are over 240,000 high-voltage transmission lines and 50 million transformers in the United States. Most large and expensive transformers are reaching the end of their life. About 70% of them have been serving for 25 years or more. Increasing loads and unstable integration of renewable energy are pushing the aging power grid to its limits.

That’s why Argonne provides operators with this asset management tool. This will help ensure the reliability and security of our electricity grid in the future. But it will also level the playing field by giving small utilities the same cutting-edge technology as large corporations.

Qiu’s team is quick to note that this research would not be possible without close collaboration with partners in the energy industry. Their long list of partners includes utilities, as well as representatives from the hydro, solar and wave energy fields, as well as academia such as Wayne State University and Iowa State University.

“Our research is a collaborative effort between scientists, engineers and industry partners,” Zhao noted. “Together we are driving positive change and shaping the future of power grid maintenance.”