Predicting Power Outages
Suddenly, everything goes dark. The light switch flicks up and down but doesn’t work. A power outage sweeps the city.
Unplanned power outages due to environmental conditions (wind, lightning, tree growth, etc.) leave those without power on their own, sometimes for long periods of time. Utility companies lack the ability to predict when forced outages will occur, so no mitigation measures targeting consumers are deployed ahead of time to reduce the impact of an outage.
Dr. Mladen Kezunovic, a professor in the Department of Electrical and Computer Engineering at Texas A&M University, and his team are combining historical outage data and weather-related data, often called big data, and machine learning to predict outages and change the outage mitigation paradigm from reactive to proactive. This will help consumers be prepared.
Using machine learning and a variety of data describing the causes of outages, the team can study data from the past to make predictions about the future.
“A lot of people ask how you test something that has never been done before,” Kezunovic said. “Say you’re making predictions using data from the past. You’re not predicting what will happen in the summer of 2024, you are predicting what actually happened in the summer of 2023, and then comparing what happened in 2023 versus what you predicted would happen. If you were correct about the past, it should work in the future.”
Once they gather the necessary data, they can superimpose database models on physics-based models to predict the state of risk of an outage.
“Wind, precipitation and lightning can all cause an outage,” Kezunovic said. “We’re dealing with over 60 different parameters like precipitation, temperature, wind, soil type, vegetation type, and animal intrusion from different databases that must be correlated. There is no human cognitive ability to correlate that manually, but machines can.”
They correlate the risk of an outage represented by a data model with the physical disposition of the transmission lines and feeders — wires that go from substations to buildings and houses using geographic information systems (GIS). GIS involves overlaying information — in this case predictions based on models — onto geographically dispersed elements such as a grid.
“If there’s rain and lighting on a specific feeder, the feeder is our physical model, then the risk of outage may occur on a section of a feeder, and because the feeder connects to the homes, these may be without electricity,” Kezunovic said.
This article by Katie Satterlee originally appeared on the College of Engineering website.