Bose proposes risk-sensitive design to address uncertainty in electric markets


Laura Schmitt

Power production from green energy resources such as wind and solar is uncertain because no one controls when the wind blows, or the sun shines. As the government pushes for more wind and solar to be integrated into the nation’s electric grids, there’s a growing need to alter grid operations to accommodate uncertainties in their energy supply.

Subhonmesh Bose
Subhonmesh Bose

The electricity markets are where supply and demand are negotiated by those entities that generate power and those that deliver it to customers. It is a highly regulated and complex process. Therefore, these markets must evolve in a way that anticipates uncertainty from renewable resources, models them explicitly, and cost-effectively operates against them.

“There are ramping constraints with certain forms of power generation, some among which could take a day to start producing,” said ECE Assistant Professor Subhonmesh Bose, referring to conventional generation methods. “The question is: how should such generators with varying degrees of ramping limitations be called in for service to tackle uncertain supply, sometimes on a very short notice, when forecasting errors are inevitable?”

Bose is addressing the uncertainty issue with $500,000 in funding he received from the National Science Foundation earlier this year through the agency’s highly coveted CAREER Awards for young faculty. In addition, he is developing a theoretical framework for a risk-sensitive electrical power supply market that can effectively accommodate the daily uncertainties of renewable power generation.

According to Bose, electricity markets currently take either a reactive or a conservative approach to uncertainty.

“One way is to take the worst-case approach; you model the worst scenario you can find yourself in and make decisions to handle that scenario,” he said. “Such a framework can be conservative and will likely lead to high operational costs. Alternately, you can try to plan for an average case; but that leaves you way too vulnerable to inevitable forecast errors.”

Bose proposes a risk-sensitive design paradigm that balances between the two extremes. His mathematical approach will combine recent advances in optimization theory, machine learning, and networked markets to enable large-scale renewable energy integration in the power grid. His framework will include risk-sensitive formulations for electricity market clearing, efficient algorithms to solve such formulations, and the design of prices to transact with the market participants.

Another component of CAREER funding is curriculum development. Bose plans to enhance his existing course instruction—ECE 598 Electricity Markets, ECE 576 Power System Dynamics & Stability, and ECE 365 Data Science & Engineering—with the latest results from his research.

The NSF CAREER Award is the agency’s most prestigious award in support of early-career faculty who have the potential to serve as academic role models in both research and education and can advance the mission of their respective department or organization.