Fengqi You Brings Energy Systems Engineering to Cornell

By Sherrie Negrea

Last fall, a report on global warming caught the attention of Fengqi You, the Roxanne E. and Michael J. Zak Professor in Energy Systems Engineering. The United Nations Intergovernmental Panel on Climate Change warned that the previous goal of limiting rising temperatures to 2 degrees Celsius by 2050 should be lowered to 1.5 degrees—or the world would face catastrophic heat waves, flooding, famine, drought and wildfires.

For You, who has researched climate change mitigation for more than a decade, the new target presented both a challenge and an opportunity. While the world’s energy systems will need to be transformed to meet the lower threshold, You believes a new tool he has developed using data-driven decision making and optimization algorithms offers the best approach to combat global warming. “Our methodology is the optimal way to make it happen,” he says confidently.

Before he arrived at Cornell in 2016, You pioneered a technique that changed the life-cycle analysis paradigm by moving it one step forward to what he calls life-cycle optimization. Instead of just measuring the environmental impact of a product, You proposed a strategy that not only determines the processes contributing the most to global warming, but also quantifies them and explores how to remediate them to reduce their greenhouse gas emissions while controlling costs.

His development of the life-cycle optimization tool while he was teaching at Northwestern University drew recognition in the world of energy systems research. And in 2015, when faculty at the Smith School of Chemical and Biomolecular Engineering started a search to fill a newly endowed professorship that would focus on energy systems, scientists around the country recommended You.

The professorship was endowed by Mike Zak ’75, general partner at Charles River Ventures, who studied operations research and foresaw the importance of bringing rigorous optimization methodologies into the emerging field of energy systems engineering. “Mike recognized that far too many analyses of energy options failed because the designer, innovator or investor accepted a myopic perspective of the problem—a single material component, a single battery cell, a single home—and failed to understand how these components interact in a system,” says Lynden Archer, the James A. Friend Family Distinguished Professor in Engineering.

Out of a field of about 200 applicants, You was selected for the endowed chair. “We believed he is exactly the person to lead the new field of energy systems engineering nationwide,” Archer says. “Before the start of our search for an inaugural occupant for the Zak Professorship, I frankly didn’t know that anyone like Fengqi existed. But fortunately I was able to consult broadly outside Cornell with persons in the know, and the vast majority of them typically ended our conversations with four words about Fengqi: ‘This is your guy.’”

While You joined the faculty in the Smith School of Chemical and Biomolecular Engineering, he is also affiliated with six other departments, from Operations Research and Information Engineering to Electrical and Computer Engineering, allowing him to work with faculty and graduate students across those disciplines. You is also a cofounder of the Ph.D. program in Systems Engineering and serves as the department’s chair of Ph.D. studies.

“One thing I really appreciate about Cornell is it’s a very open system,” he says. “We can easily collaborate with people across departments. There are no boundaries, and this is fantastic because it’s much easier for cross-disciplinary research collaboration.”  

Just two years after arriving on campus, You received the Cornell Engineering Research Excellence Award in 2018. The award recognized his innovative research in computational models, optimization algorithms, and machine learning as well as his impressive record of publishing about 25 papers a year.

You also won four international awards from professional societies over the past two years for his research on sustainability, energy and computing, while also receiving a number of best paper awards.

“Fengqi is unusual for a faculty member at any stage of their career in terms of his activity level and engagement with many projects across many disciplines,” says Abraham Stroock, the William C. Hooey Director of Chemical and Biomolecular Engineering and the Gordon L. Dibble ’50 Professor. “He just has more hours in the day, but it’s characteristic of his discipline in systems engineering. He can bring those tools to any discipline or problem, and he does.”

One university-wide project You has worked on is the Cornell Initiative for Digital Agriculture, which is harnessing technology and data systems to optimize all aspects of food production. Collaborating with Stroock, You has applied machine learning and optimization to synthesize data from weather forecasting, marketing conditions and regulatory information to help farmers improve their operations.

In one project, You has used this technology to create a system for smart irrigation. If sensors on a field, for example, show that crops are dry, a farmer may still delay watering after checking the system and learning that a heavy storm is approaching.

“Traditionally, people know what’s going on now and what happened in the past,” You says. “They don’t know about the future. Our answer is we can look at the future by leveraging big data.”

Agriculture is just one of many fields You’s work has impacted as it has transitioned to incorporating artificial intelligence into optimization algorithms. A few years ago, he coined the term “grey box-digital twins” to describe this approach, which combines data-based (black box) and model-based (white box) methods.

One company that has adopted You’s approaches is The Dow Chemical Company, which is working to integrate his method of stochastic optimization for manufacturing production. Historically, the company plans production based on an average of forecasted demand, but this does not take into account the errors in the forecast and uncertainty in demand, says John Wassick, a Technology Fellow for Integrated Supply Chain at Dow Chemical in Midland, Michigan.

“We produce products ahead of orders, but we have to swing between different products, and how much of any product we make is subject to uncertainty,” says Wassick, who served on You’s dissertation committee at Carnegie Mellon. “What Fengqi was able to do was show how to use the stochastic optimization methods that would formally take into account the uncertainty in future demand.”

In a simulation of the optimization tool, the company concluded it could save 5 percent in product costs, which would translate into millions of dollars for some of Dow’s businesses. Wassick said Dow is now looking for the optimal circumstances to implement the method.

Another company that is using You’s research is ExxonMobil, which is working to reduce its environmental impact by lowering carbon emissions and increasing efficiency in its oil refining and petrochemical production, says Robert Johnson, section head, modeling and optimization, in the Process Technology Department at the company’s research facility in Annandale, New Jersey.

“I think what the work of Professor You is doing is really helping us take that to the next level—rather than simply assessing impact, this work actually let’s us make technology decisions based on the potential impacts on greenhouse emissions and the carbon footprint,” Johnson says.

Johnson says You is one of a few researchers in the world who are combining the techniques of life cycle assessment and mathematical optimization. Because that research aligns with the work being conducted at ExxonMobil, Johnson has hired several of You’s Ph.D. and postdoc students.

“His research group is somewhat unique in that he combines sustainability research with rigorous mathematical optimization,” Johnson says. “And the people who do their Ph.D.s with him come out with very unique skills that are difficult to find.”

You’s work on sustainability has challenged the potential of renewable energy sources in unexpected ways. A study he co-authored in 2015 showed that silicon-based solar panels have a more detrimental ecological impact than modules made from a new class of materials called perovskites. His study showed that while the energy payback time for silicon panels was two years, it was two to three months for modules made with perovskites.

At Cornell, You recently concluded a study showing that energy usage in buildings can be reduced by up to 10 percent by using a data-driven system predicting the accuracy of weather forecasts. The system, which relies on years of data on weather forecasts, actual weather conditions, and building characteristics, was tested on the Toboggan Lodge, a 90-year-old building on campus.

“This is a smart energy system that combines data-driven decision making, machine learning, automatic control, and optimization to optimize the energy system,” You says.

Another campus initiative You has helped lead is the Cornell Energy Systems Institute (CESI), which was restructured last year after Archer was named its director. As part of the reorganization, the institute was renamed by adding the word “systems” to its title.

“The importance of that one word cannot be overstated—it changed almost everything about our approach, mission and ambitions as an institute,” Archer says. “It was clear to us that the knotty interactions between components so often ignored by others provide fertile ground for Cornell to differentiate itself from our peers. Once the decision was made that systems approaches would guide all of CESI’s work, it was kind of natural to recruit Fengqi as a key member of the institute’s leadership.”

The institute will focus on translating research into technology that will lower costs, improve performance, and reduce the carbon footprint of energy systems. One of the institute’s four associate directors, You also serves as the faculty advisor for the new student-led Energy Systems Club, which will work with companies to solve their energy problems.

Looking toward the future, You predicts that reducing greenhouse gas emissions will require a hybrid energy system that moves beyond a single solution. One prime example is Cornell’s initiative to become carbon neutral by 2035, which, as You described in a recent paper, would be possible through a combination of lake source cooling and adoption of renewable energy sources such as geothermal and biomass.

That transformation, however, would increase costs by 5 percent over the university’s current spending on energy, which relies on natural gas. “Nothing is free,” You says. “Natural gas is very cheap in the U.S. If you are going to move to more renewables, there’s a penalty with that. So there’s always a tradeoff. That’s what optimization is all about—determining the tradeoffs, the upside and the downside, and picking the better alternative for a sustainable future.”