A senior capstone project sponsored by Ohio’s Butler Rural Electric Cooperative at a local university is giving the co-op insight in how to use artificial intelligence to efficiently manage rolling outages—while also providing students real-life experience with a critical issue facing utilities.
Working with the College of Engineering and Computing at Miami University, the Oxford-based co-op asked students to create an interface to provide real-time data to help the operator of its supervisory control and data acquisition (SCADA) system manage rolling outages and document their impact on the system.
While the co-op isn’t currently part of any load shed plans, they figure it can’t hurt to plan ahead since “it might be a possibility in the future,” said Cody Holthaus, Butler Rural Electric Cooperative’s manager of engineering and project mentor. “We decided to go ahead and start making efforts to put together a load-shedding plan to roll out in case we do need to react.”
Using data from the co-op, the students used AI to simulate a load monitoring environment in which an operator could decide which feeders to energize or de-energize and when.
“The goal is to ensure that non-critical members evenly share the burden or downtime associated with rolling outages,” Holthaus said.
Cost and proprietary issues involving the developer of the co-op’s SCADA system prevented the project from being fully deployed, and the final result was a standalone application. But preliminary results can help the co-op track effects of some rolling outages, Holthaus said.
Located in the co-op’s service area, Miami University each year asks local businesses to sponsor capstone projects. Sponsors propose ideas and, if selected by students, they commit to working with students for an academic year. Sponsors also donate funds and provide technical assistance.
“Butler is a valuable partner,” said Husan Ul Banna, a visiting assistant professor who plans to build on the project for another senior capstone next year.
The project, “Rolling Blackouts and Load Shedding—Machine Learning Optimizations,” won national recognition as well. Earlier this year, one of the students, Owen Campbell, presented his team’s findings at the annual Institute of Electrical and Electronics Engineers Power and Energy Society conference in Anaheim, California, where it won first place in the undergraduate student poster competition. The other team members were Shaun Valentine and Mina Yang.
Now a substation engineer at POWER Engineers in Columbus, Campbell said the team took on the project because they were “interested in applying AI in real-world solutions, especially in the power engineering field. The opportunity to work with a local company was very appealing.”
Victoria A. Rocha is a staff writer for NRECA.