UMass Amherst Faculty, Alumni Receive Test of Time Award for Sustainability Research

A team of faculty and alumni of the University of Massachusetts Amherst, including Distinguished Professor Emeritus Jim Kurose and Distinguished Professor Prashant Shenoy of the Manning College of Information and Computer Sciences (CICS), and Professor David Irwin of the College of Engineering, received a Test of Time Award at e-Energy 2024, the conference on future and sustainable energy systems hosted by the Association for Computing Machinery this month in Singapore. 

Along with lead author and alumnus Aditya Mishra ‘15PhD and Ting Zhu of Binghamton University, they were honored for their e-Energy 2012 paper, “SmartCharge: Cutting the Electricity Bill in Smart Homes with Energy Storage.” This early paper on time-shifting energy usage established what is now a popular approach for decarbonizing the residential electricity grid, the cloud, and other settings. 

Mishra, an associate teaching professor at Northeastern University, is a former doctoral advisee of Shenoy and alumnus of his Laboratory for Advanced System Software (LASS), which operates with the goal of learning to use computing technology to green our infrastructure and reduce carbon emissions while aiming towards a more just society. 

Responding to new electricity pricing models being rolled out around the country at that time, the team’s paper aimed to solve this question: how can market-based solutions that reward energy usage at off-peak times serve households that cannot always choose when to use electricity?  

“If you work during the day, you may not have the flexibility to move the time you do your laundry or dishes to a low-price time of the day,” Shenoy says. “So, what do you do? You are stuck with a higher electricity bill.” 

The answer: a smart battery that draws energy from the grid during low-price periods and discharges to the household during high-price periods. The team’s approach, called SmartCharge, proposed an intelligent charging system that could forecast future demand using statistical machine learning techniques, and make decisions based on next-day electricity markets about the best times to draw from the grid. 

The idea has since been widely adopted, finding direct use in products such as energy storage products and in offerings from utility companies that give customers a discount on energy use in exchange for allowing their smart battery to be managed by the utility company in a way that helps manage peak demand on the grid. This approach by utility companies, known as grid peak shaving, brings a specific focus to reducing the peaks of energy use across the grid. 

“During peak periods, you need to have excess generation capacity that's brought online so you can serve the extra demand,” says Shenoy. “These are going to be your least efficient generators, and they’re usually not very clean in terms of emissions.” 

As Shenoy explains, reducing peak demand across the grid reduces the use of these less efficient and less clean generators. Because electrical grids must always be able to meet the full demand, reducing peaks also reduces the need to build excess generators and the capital costs that are passed onto consumers. 

At the time of publication, the team demonstrated that their smart battery approach could reduce household electricity costs by 10–15% and, if widely deployed, would reduce grid demand peaks by 20%.