Skip to Main Content

Energy is one of the most valuable resources in the world, and we depend on it for our daily lives. Power grid networks are the main source for electricity delivery; without power grids, society loses electricity, with natural gas and water failing soon after. Much like anything else, these grids are subject to disturbances and errors, leading to costly problems needing fast solutions. To find these solutions, Aayushya Agarwal and his team, Amritanshu Pandey and Larry Pileggi, introduced an algorithm to quickly facilitate highly accurate simulations and optimizations of these grids, ensuring energy is never found in short supply.

The power grid in Texas failed in February 2021, leaving millions of people without power in the middle of unexpected severe winter storms. “The Texas grid is a worst-case scenario,” says Agarwal, a Ph.D. student in electrical and computer engineering. “A lot of engineers are trying to prevent such situations by doing an analysis beforehand and seeing what kind of corrective measures they could take to prevent these failures.”

These analyses, however, have some shortcomings that they seek to improve. Currently, methods to locate disturbances in a grid involve reworking their entire mathematical problem from the beginning, leading to longer–and therefore more costly–waiting periods before a solution can be found to restore a power grid.

The grid is a very dynamic system that requires constant observance.

Aayushya Agarwal, Ph.D. student, Electrical and Computer Engineering

Instead, methodology crafted by Agarwal and his team uses prior information about the grid as a starting point to find the solution, efficiently translating a previously known network configuration into a new one without the disturbance.

What disturbances can affect a power grid? According to Agarwal, they can be simple, such as a line going down or a generator being switched off. However, more complex issues, such as severe weather, require even faster solutions. “The grid itself is a very dynamic system that requires constant observance,” explains Agarwal.

“As a result, we need our analyses to be very fast. Power grid engineers are trying to analyze all of these aspects beforehand, to ensure that their grid is actually stable.”

The challenges are two-fold: finding a solution and implementing it in a feasible amount of time. Their research on simulation and optimization will affect both industry and academia, having a place in the operational engineering used to monitor the energy grids, as well as pushing the boundaries of “what-if” scenarios in other forms of research, including large-scale transmission resiliency and planning studies.

Through these studies, the new methodology has the potential to solve problems before they ever occur, saving valuable time and money.