Could AI protect renewable energy sources on the grid?
Image credit: City of Ames
Iowa State engineers are working to secure renewable energy sources, such as solar and wind farms, against cyber attacks.
According to experts, solar panels and wind turbines, now projected to produce 44 per cent of America’s electricity by 2050, present cyber-security challenges.
Many of these renewables have sensors, controllers, actuators or inverters that are directly or indirectly connected to the internet, and they’re distributed far and wide across the country and the countryside.
Many have insecure connectivity to legacy electric grid systems, therefore making them subject to advanced persistent threats. More of these systems will also be online over the next few years.
So there’s a need for cyber-security systems that “prevent, detect and mitigate” attacks on renewable sources sending power to the grid, said Gelli Ravikumar, a research assistant professor of electrical and computer engineering at Iowa State University.
Ravikumar is leading a team that is developing such systems for protection as part of a $12m, six-project effort by the US Department of Energy.
“Investing in cutting-edge cyber-security technology keeps us at the forefront of global innovation and protects America’s power grid in the face of increasing cyber threats from abroad,” said Jennifer Granholm, the US secretary of energy, when they announced the grants this spring.
The energy department is supporting the Iowa State-led project – penned 'GridDeep' – with a three-year, $2m grant which will see the team research, develop and demonstrate a physics-based system that uses artificial intelligence (AI) tools to improve the cyber security of the country’s energy-delivery systems.
“The project’s next-generation, AI-integrated cyber-physical security technology, and tools can aid in ensuring energy-delivery systems are designed, installed, operated, and maintained to quickly recover from cyber attacks,” Ravikumar said.
Because those energy-delivery systems are complex and constantly changing, Ravikumar said AI technology will help electricity producers better understand their systems, quickly respond to attacks and restore critical functions.
The project is especially timely because “a lot of green energy is being integrated into the grid”, said Manimaran Govindarasu, an Iowa State Anson Marston distinguished professor in Engineering, and a co-leader of the research team. “We want to ensure the security and resilience of this renewable integration.”
According to Govindarasu, a major challenge is that these renewable power sources operate more like smaller municipal or co-op power plants. There are also limited resources for protecting them and more opportunities for attacking them.
Govindarasu stressed the need for innovative solutions to tackle this, with the potential to scale up.
The Iowa State engineers said it will take a group effort to secure the growing number of renewable power sources – and so the project includes researchers from Iowa State, the University of Texas at El Paso, the Electric Power Research Institute, Duke Energy, Alliant Energy, Google, OSIsoft and Schweitzer Engineering Laboratories.
“We’re bringing together IT companies, energy-system vendors, and utility companies,” Govindarasu said. “This will be an opportunity to use their expertise and advance this technology. There will be more synergy, collaboration, and innovation.”
This will lead to “cyber-physical security tools and technologies that are not yet available but can become widely adopted throughout the energy sector to reduce the risk that a cyber incident could disrupt energy delivery systems”, Ravikumar added.
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.