Researchers at the Department of Energy’s Oak Ridge National Laboratory are utilizing sensors, drones, and machine learning to prevent fires and lessen the damage they cause to the electric grid as a result of larger and more frequent wildfires caused by climate change.
Engineers are developing technologies to detect electrical arcing, broken equipment, and the direction of growing flames remotely.
The need is immediate. Several of the warmest years on record have coincided with the years with the most wildfires since 2004, according to statistics from the National Interagency Fire Center. Electric utilities have a big stake in finding answers to the issue since many wildfires are ignited by power lines and because they can cause prolonged power disruptions.
The DOE targeted funding for a few wildfire programs during the previous fire season, including two at ORNL.
We chose this research to be accelerated because we recognize how critical and hazardous wildfire has become due to climate change and other factors.
Stewart Cedres, Technical Lead and Strategist, Department of Energy
According to him, transmission line damage from wildfires can result in brownouts and blackouts many states away.
Cedres added, “The work ORNL is doing with sensor technology on drones that can go beyond line of sight is unique and very critical because it allows us to be several steps ahead of issues that can put the grid at risk, making it unreliable and less resilient.”
Cedres claimed that DOE has two interests: preventing electric grid fires from starting and defending the grid from fires. Both of these issues are being worked on by an ORNL team under the direction of Peter Fuhr, head of ORNL’s Grid Communications and Security Group.
The U.S. Forest Service, which is the main organization responsible for battling wildfires, is also strengthening its fire detection sensors.
When outlining how the Biden Administration is combating the threat of wildfires, the White House cited the wildfire study being undertaken by Fuhr’s team in July 2022.
Power flow fluctuations that result in superheated sparks can be detected by sensors installed close to power lines or inside power electronics equipment. To quickly identify and categorize these electrical arcing events—which happen when electricity leaps through the air between two conductors—ORNL researcher Ali Ekti is working on an algorithm.
Utility companies and academic institutions have provided a library of waveform “signatures” that Ekti is utilizing to display voltage or current spikes brought on by anomalies like arcing. These anomalous zones are discovered by his algorithm.
The next stage will be to teach a computer algorithm to identify signatures that are similar and locate the position of the arcing. The method would notify the utility command center so employees could examine the site to see whether it is accessible and if it could be implemented into power electronics equipment.
Here is where drones can help. Arcing frequently takes place in remote, difficult-to-access terrain. Sometimes it is difficult for helicopters and planes to go low enough or near enough to spot tiny arcs. Additionally, Fuhr noted that they are more expensive to fly and challenging to plan rapidly.
Using drones with sensors that can film video and detect noise, heat, and radio frequency emissions from tiny arcs is a possible substitute. These merged sensory abilities can validate issues that have been found and eliminate interference.
Prior to a fire starting, drones can detect maintenance issues. Transmission lines can be inspected over considerable distances due to ORNL’s expertise in using drones outside of the visual line of sight.
Fuhr’s team recently received unique clearance from the Federal Aviation Administration to fly drones throughout the country and up to 1,500 feet with special approval.
Drones are already out there doing vegetation management. So why don’t we use an additional technology on the same platform to check for arcing at the same time? We are establishing demonstrations in the coming year at utilities large and small across the nation.
Peter Fuhr, Head, Grid Communications and Security Group, Oak Ridge National Laboratory
It can be difficult for drone-mounted audio sensors to differentiate between the drone propeller’s buzz and the arcing’s crackling noise. However, Ekti has created a new method that uses math to separate the arcing audio signature from background noise.
Directing Fire Fighting
At another presentation for the Forest Service in October, Fuhr’s team might put this to the test in the field. ORNL researchers will demonstrate how they can utilize drones to thermally identify sparks before a significant fire in the Kootenai National Forest in Montana.
This summer at a neighboring farm, the crew put the method to the test by starting a confined campfire amidst strewn little gel fuel cans. They conducted tests to determine the altitude at which a sensor installed on a drone might detect the small fires below.
According to Marva Willey, a U.S. Forest Service fuel specialist who is a member of the agency’s tools and technology team, this capability might be a game-changer in the battle against wildland fire.
Heat-detecting sensors have so many applications that can increase our situational awareness and knowledge of where heat and fire are. It is all about knowing what is going on before you put people out there.
Marva Willey, Fuel Specialist, U.S. Forest Service
According to Willey, sensors attached to drones could be able to detect new flames started by embers that can spread up to a mile from the initial fire.
The drones might also be used to find any hot spots still present in a burned-over region, something that firefighters now do by hand while sorting through ash.
Researchers from ORNL are supplying sensors for a Forest Service project involving the detection of heat at extremely high elevations. Sensors might be carried by a balloon or a long-range solar-powered aircraft flying above thunderstorms to look for fires started by lightning strikes.
Later this year, testing of the technology is planned. According to Willey, it might lessen the need for piloted fire-spotting flights, reducing the risk to people while making judicious use of the few resources.
Fuhr’s team has started looking into further ways to avoid electrical fire using drone-mounted sensors, such as splice or transformer heat detection.
In 2021, Fuhr and his team were able to utilize drones to detect heat in transmission lines and transformers while a wildfire sparked by downed power lines was raging in the neighborhood of Wears Valley, Tennessee.
Transformers might be damaged by exposure to temperatures that are greater than their operational limitations, even if they survive the fire, according to Fuhr. This could point to where equipment needs to be changed.
Fuhr added, “New Mexico was on fire earlier this year, and when you see the devastation that happened with the (2018) Camp Fire and the resulting bankruptcy of the utility—if our activities can reduce those kinds of outcomes, it is good for everyone.”