Margaret Bones had to wait for the rotor blades on the Bell Long Ranger helicopter to begin to whirr before there was electricity to power up her computer, cameras and sensors, but once the blades were spinning, she punched the buttons turning on her gear.
Bones and her pilot Raphel Perozo were headed to Indian Hills, a neighborhood east of Evergreen in search of wildfire risks using high resolution cameras and laser radar, known as lidar.
The search has taken on heightened urgency in the wake of the 2021 Marshall Fire, the most destructive in Colorado history razing more than 1,000 Boulder County homes and businesses, tallying $2 billion in damages.
One cause of the fire, county officials concluded, was an Xcel Energy distribution line that had come loose in high winds. The utility has disputed that finding, but is already facing lawsuits from homeowners and businesses.
Still, electrical lines do pose a risk, the company noted in a filing with the Colorado Public Utilities commission, citing a 2019 study by Verisk Analytics that found one-third of the more than 2.2 million housing units in the state are in either moderate or high wildfire risk areas.
That ranked Colorado third behind California and Texas for properties exposed to wildfire danger.
And so, the hunt is on.
The use of lidar radar and computer analysis of the data it amasses are part of Xcel Energy’s Wildfire Mitigation Plan, submitted to state regulators in 2019 and funded with $148 million in ratepayer dollars.
Bones, 27, an airborne systems operator for Sandy Springs, Georgia-based GeoDigital, has flown lidar runs all over the U.S. and Canada, sometimes collecting data for self-driving auto systems and sometimes for utilities checking their high-voltage transmission lines.
More and more of the work is checking smaller power lines, Bones said.
“LiDAR is more commonly applied to transmission lines,” Drew McGuire, director of research and development at the nonprofit Electric Power Research Institute, said in an email. “With increasing focus on wildfire risk reduction, some utilities are finding ways to use this technology on distribution lines as well.”
McGuire said that these applications are emerging technologies. “Not all LiDAR is created equal,” he said. “They’re easy to use, but tricky to get ‘accurate’ data.”
Still, the experience with lidar is rising as the costs of the technology and data analysis decline. “New innovations to collect and process these data will likely make LiDAR application in distribution even more common in the future,” McGuire said.
Bones and Perozo are searching for possible problems with vegetation, inadequate clearance for lines and potentially overloaded poles.
The project, begun in 2020, aims to cover 1,500 miles of distribution lines in more than 300 wildfire risk zones by 2025. Last spring — spring is the prime time to do this work — the project scanned the San Luis Valley. This year it is scouring the foothills west of Denver.
The work is painstaking as it can take multiple passes to complete one serpentine line. “On a good day we can do 20 miles of line, but usually only manage 12 to 15 due to afternoon storms,” Bones said.
On the belly of the Long Ranger — a bigger, more powerful version of the Bell Jet Ranger to deal with mountain turbulence — is the lidar sensor, three high resolution cameras and a near-infrared camera. There is also a weather probe recording wind, temperature and humidity.
As they trace a line, Bones watches the images from the three cameras — one facing forward, one facing down and one facing aft — on her computer monitor. Once the line has cleared the rear camera, Bones knows the section is done.
At the same time the lidar — short for Light Detection and Ranging — is sending out laser pulses measuring distances and generating data that is the basis for a 3D representation of the lines and poles. The near-infrared camera can also identify unhealthy trees near lines.
All this data — up to four terabytes each day or the equivalent of uploading one million photos — is then used to create a 3D computer model of the distribution system by an Xcel Energy contractor, Fort Collins-based EDM International.
With the model the network can be stress tested for temperature, wind loads, and to see if lines are adequately spaced and whether poles are overloaded.
“We are looking for those inadequate clearances and overloaded poles,” said Andy Stewart, president of EDM International.
Finding potentially problematic poles without some kind of scanning would be a challenge in as much as Stewart estimated there are 89,000 poles in wildfire risk zones.
The data fits into other Xcel Energy initiatives, such as its Mountain Hazard Tree Program focused on addressing tree mortality from pine bark beetles.
The lines can be tested against simulated 90-mph winds. Would that catch a line being blown loose by the wind?
“Yes, LiDAR modeling enables Xcel Energy to identify poles within the wildfire zone that could become overloaded due to strong winds,” the company said in a statement. “When these poles are identified, Xcel rebuilds these poles to Grade B Heavy loading of the National Electric Safety Code (NESC), which is the most conservative standard.”