Numerous projects are underway in the Midwest to prevent bats and birds from colliding with wind turbines. But little reliable information is available about interactions between birds and large-scale solar installations. Scientists at Argonne National Laboratory in Illinois are working to change that.
“It was one of the biggest mysteries for the solar photovoltaic developers and operators,” said Yuki Hamada, Argonne scientist and project lead. “We wanted to help develop the technology to really solve the mystery of avian-solar collisions.”
Birds are difficult to observe in real time, so little is known about how they actually use solar sites, Hamada said. That creates uncertainty about best practices for mitigating interactions, and the research could also reveal how birds might benefit from the presence of solar panels.
The Argonne team will create and install artificial intelligence-enabled advanced cameras to monitor bird activity at several different solar sites, including one operated by utility Duke Energy. The cameras will gather data about what happens when birds fly past, perch on, or collide with solar panels. The scientists anticipate that learning more about how and why collisions and bird deaths occur can help to prevent them.
The team is using AI-enabled “deep learning” to build models that train computers how to distinguish birds from other animals or objects. The computers are also being trained to identify specific bird behaviors.
Without high-tech cameras, a lot of human field work is needed to track bird deaths and figure out what causes them. But wildlife behavior changes the more humans are present. Cameras reduce the human effect, collect more accurate and complete data, and lower surveillance costs.
Newer iterations of the Argonne technology go beyond the original purpose of preventing collisions. Monitoring birds’ activity around solar facilities could provide ecologically meaningful information about their behavior, Hamada said, which suggests how solar installations could actually support birds and other wildlife.
For example, initial observations suggest birds might use solar fields for nesting. Being equipped with greater knowledge of this phenomenon would help solar developers foster a more inviting and safer habitat.
“Our system is very small and very flexible,” Hamada said. “The system can be installed at different angles and different distances so that the bird movement can be observed… in multiple ways to help modelers, scientists, and engineers to better understand the behavior.”
The U.S. Department of Energy provided $1.3 million in funding for the three-year project, which will continue through the end of next year.
The technology is still under development. At this stage, the researchers are testing their lab-generated algorithms in cameras to ensure correct performance in the field, Hamada said. Then, the scientists will test the system at partner facilities. They anticipate a working prototype will be ready for demonstration next spring.
Argonne is exploring multiple options for getting this technology into widespread use in the solar industry, Hamada said. They’re examining commercialization methods including software licensing and releasing a hardware-software package.
The scientists also are investigating the system’s potential to learn about avian interactions with other clean energy technologies, like wind turbines.
“The technology may be helpful to understand the impacts of the introduction of offshore wind, which is very difficult to do because there is no such thing as a field survey,” Hamada said.
Meantime, the team emphasizes that although tech development often focuses on problems – like avian collisions – they focus on the positive side, such as providing solutions to improve wildlife while producing clean energy.
“There’s not many technologies or developments focused on the positive side,” Hamada said. “We want our technology to help that part of solar development.”
Photo: Barn swallows compete for a perch atop a solar panel in California, by Don McCullough/Creative Commons