Full Name
Sivaranjani Seetharaman
Company
Developed models to evaluate how electricity systems will respond to soaring levels of demand
Brief Biography
Our aging grids face growing strains: Renewables generate a rising share of power but fluctuate wildly with the weather or time of day. Electric vehicles demand ever more power. And increasingly extreme heat waves, floods, and storms continually threaten to knock electricity systems offline.
Making matters even more complicated, more homes and businesses are equipped with their own solar panels, batteries, or microgrids, becoming sources of electricity as well as consumers of it.
Sivaranjani Seetharaman, 33, an assistant professor of industrial engineering at Purdue University, is developing tools to keep our grids running reliably in the face of these challenges.
Among other efforts, she has developed models to evaluate how electricity systems will respond to soaring levels of demand and extreme weather, depending on the mix of sources, storage, and other infrastructure. In one instance, she and her colleagues found that if Texas’s heavy trucking sector went electric, only 11% of the fleet would have to charge at once to risk destabilizing the regional grid.
Seetharaman has used machine learning and her models to train algorithms that can help grid operators manage these increasingly dynamic, complex systems. The software tools can forecast supply and demand, or help to determine the optimal electricity sources and ideal pathways along transmission and distribution networks at any given moment, as demand, supply, and weather conditions shift.
They can also help incorporate what are known as demand response systems, through which grid operators encourage customers to dial down their energy use during moments of peak demand or, under certain conditions, do it on their behalf.
Fully addressing the grid’s looming challenges will require adding vast amounts of clean power generation, energy storage facilities, and hardware over the next few decades. But developing better algorithms and other software tools can rapidly improve the performance of today’s grids, Seetharaman says, and ensure that we’re building more efficient, flexible, and robust ones for the future.
Making matters even more complicated, more homes and businesses are equipped with their own solar panels, batteries, or microgrids, becoming sources of electricity as well as consumers of it.
Sivaranjani Seetharaman, 33, an assistant professor of industrial engineering at Purdue University, is developing tools to keep our grids running reliably in the face of these challenges.
Among other efforts, she has developed models to evaluate how electricity systems will respond to soaring levels of demand and extreme weather, depending on the mix of sources, storage, and other infrastructure. In one instance, she and her colleagues found that if Texas’s heavy trucking sector went electric, only 11% of the fleet would have to charge at once to risk destabilizing the regional grid.
Seetharaman has used machine learning and her models to train algorithms that can help grid operators manage these increasingly dynamic, complex systems. The software tools can forecast supply and demand, or help to determine the optimal electricity sources and ideal pathways along transmission and distribution networks at any given moment, as demand, supply, and weather conditions shift.
They can also help incorporate what are known as demand response systems, through which grid operators encourage customers to dial down their energy use during moments of peak demand or, under certain conditions, do it on their behalf.
Fully addressing the grid’s looming challenges will require adding vast amounts of clean power generation, energy storage facilities, and hardware over the next few decades. But developing better algorithms and other software tools can rapidly improve the performance of today’s grids, Seetharaman says, and ensure that we’re building more efficient, flexible, and robust ones for the future.
Speaking At
