Light a Light or Light a Fire

cover photo from iStock

written by Xinyue Wang

When people think of California, they usually picture the generous sunshine, gorgeous beaches, and numerous landmarks. However, nowadays the image probably has gone a bit downhill as ‘wildfires’ and ‘power shutoff’ drop frequently in the media headlines. While it’s just heartbreaking to see the Golden State burning so wildly, many of us are also confounded — What caused the wildfires? Why California? How can this be stopped? And most confusingly, what does the power shutoff have to do with the wildfires?

In our study, recently published in Nature – Scientific Reports, we explored the problem of electricity power infrastructure-related wildfires and proposed a  methodology for predicting wildfire ignition caused by powerlines (i.e., conductors) coming in contact with vegetation. 

A recipe for fires

As with many other natural hazards, wildfires exhibit regional characteristics and local conditions affect typical ignition causes and fire behaviors. While wildfires can be started from various sources (e.g., lightning, smoking, etc.), electrical powerlines were shown to be a significant ignition source in California. Statistics showed that out of the top 20 most destructive California wildfires, power systems caused at least five of them.  

A power network, consisting of numerous components and equipment, can experience a sharp increase in failure or faults under strong winds. In high-voltage transmission systems, the transmission conductors are easy victims as they span long distances across variable terrains and are highly flexible and can swing up to 20 meters in the middle of the span.  

This movement is dangerous because flashover can occur when the conductor gets close to nearby vegetation, typically trees. Flashover is a phenomenon where the electrical current jumps through air from the conductor to a nearby object. The energy released from the high-voltage current can result in ignition and even fires if combustible fuels are present in the near vicinity. Did we just mention the trees? 

The winds don’t escape from the crime scene. Instead, they stay to facilitate the fire spread while also hindering fire fighting efforts. You have probably heard of the intense foehn winds (known as Diablo winds or Santa Ana winds in California) from news reports on wildfires. 

Now the recipe for fires seems clear: wind, conductor, vegetation, and maybe hot, dry air (the easiest to get in California). We must mention that this recipe for fires is subject to change as climate change is spicing things up. 

A trolley problem

Now we’ve seen where the power systems come into play in wildfires. You may have guessed why power shutoffs often happen during fire seasons. California electric utilities are authorized to shut down the power systems preemptively in response to severe weather conditions, which is called Public Safety Power Shutoff (PSPS). Despite the immediate effectiveness in stopping power infrastructure from causing fires, a PSPS event can lead to other significant disruptions. Imagine driving without traffic signals, or a patient who relies on a powered ventilator.  

The Principal Investigator and founder of Lehigh’s Catastrophe Modeling Center, Dr. Paolo Bocchini, describes this decision making as a classical “trolley problem“ in ethics, “If they do not use the PSPS, a fire may start and some people may die and enormous damages to properties and the environment may be caused. However, if they enact the PSPS, the absence of electricity for personal medical devices, general illumination, etc. may cause other serious issues and potential deaths, in addition to all the implications for businesses.”

Such decision making certainly cannot come from gut instinct but must be informed with solid predictions. This is where probabilistic risk analysis shines. In the context of PSPS, the two risks we need to compare on the scale are shown in the figure above. The scale — balancing two decisions that have respective risks — lead us to a decision that has less risk. But either decision actually has its own problem (risk).

Ignition prediction

Our study focused on ignitions caused by powerlines (i.e., conductors) swinging out to nearby vegetation under high winds. The prediction of ignition has great influence on wildfire risk analysis because subsequent fire propagation simulation and fire damage analysis rely on ignition location and timing as input.

Nowadays, predictions are popularly made by Artificial Intelligence (AI) using data-driven methods. Our study is the first of its kind applying a rigorous probabilistic approach to the problem which includes consideration of the mechanical behavior of the conductor cables under strong wind. This mechanistic approach enables a superior level of understanding and insight that is crucial for decision making.

Winds are gusty and uncertain, as are the conductor dynamic displacements. We found that it is important to consider the wind turbulence-induced dynamic effects and associated uncertainties when determining a safe clearance between the conductor and the trees. If you are interested, you can read more in our article, Predicting Wildfire Ignition Induced by Dynamic Conductor Swaying Under Strong Winds.  

That said, you shouldn’t give up your plan to visit California as long as it’s not ‘on fire’.