The Human Variable: How Shuchen Cong Teaches Energy Systems Modeling

The energy transition is not just an engineering challenge. It is a policy challenge, a finance challenge, a community challenge—and increasingly, a data challenge. The professionals who will navigate it successfully are those who can move fluidly between technical models and human realities, translating complex analysis into decisions that affect real people and places. That is precisely the kind of thinking Shuchen Cong, PhD, cultivates in her classroom.

Cong is a Lecturer in Dartmouth's Master of Energy Transition program, a graduate degree designed to prepare students from a wide range of disciplines to lead across sectors at a pivotal moment in the history of energy. Her course, Energy Systems Modeling, introduces the computational tools and analytical frameworks that underpin modern energy decision-making—from power grid planning to transportation, housing, and beyond. But what sets the course apart is the perspective she brings to it.

Research Rooted in Community

headshot-shuchen-cong-400x425.jpg

Shuchen Cong, Lecturer and Postdoctoral Research Fellow.

Cong's path to the classroom runs through years of research bridging individual energy behavior and large-scale systems—work that consistently asks how energy moves through people's lives.

Before joining Dartmouth, she was a Policy Researcher at the National Renewable Energy Laboratory (now National Laboratory of the Rockies–NLR), where she worked across a wide portfolio: distributed energy systems, building performance standards, market segmentation, cybersecurity, and more. One project brought her to Atlanta repeatedly to survey local residents and business owners about how they used their buildings and which ones their communities valued most. The findings were delivered to the local government to help prioritize which buildings to retrofit for energy efficiency.

"The Atlanta experience taught me how much is possible when you know who to ask, what to ask, and how to translate findings for specific groups of stakeholders," Cong says. "That is what makes a good designer and interpreter of models."

It is a lesson she carries directly into the classroom. Her research has always started from the demand side—from households and communities—rather than from the grid outward. That orientation shapes the examples she uses in class and the kinds of questions she encourages students to ask.

Flexibility is built into the course structure to meet students where they are. Because students come from such diverse backgrounds, they can choose among assignment options that best fit their goals and career trajectories, and they are encouraged to revise and resubmit work without penalty. "Revising and resubmitting assignments lowers the barrier to learning, and takes the pressure off of making mistakes in assignments," Cong says. The approach sometimes places more grading burden on the instructor, but the payoff in student growth is evident and worthwhile.

A Course Built for a Wide Range

Energy Systems Modeling is deliberately broad—asking students to think across sectors rather than within any single one. In an industry where the decisions of a utility planner, a municipal policymaker, and a private equity analyst can all converge on the same infrastructure project, that cross-sector fluency matters.

Cong structures the course in two halves. The first introduces students to foundational energy system models and individual pieces of the energy system—power systems, demand and load growth, fossil fuels, renewable systems, a guest lecture on cybersecurity—building familiarity with the landscape. The second half shifts to specific modeling methods and best practices, drawing on data from across those sectors and asking students to work with it in integrated ways. Students noted that the structure and pacing worked well; themes that appeared early kept resurfacing, reinforcing their understanding with each new context.

The course is designed not for technical mastery of model building, but for analytical skill and confidence. "Because the course is short and serves such a wide range of students, the goal is to build the vocabulary, logic, and general rules to assess and critique models for different stakeholders," Cong says. That means learning to examine a model's inputs, outputs, and the methods that connect them—and knowing where to dig when something doesn't add up.

Learning to Work with Uncertainty

inline-met-class-600x400.jpg

MET students complete three terms, with three courses held each term. (Photo by Beam Lertbunnaphongs '25)

If there is one concept Cong most wants students to carry into their careers, it is a deep respect for uncertainty. Energy systems are full of it. Technology costs shift. Policies change. Human behavior resists prediction. And the models that guide billion-dollar infrastructure decisions are built on assumptions about all of the above.

"Uncertainty is critical to understand," she says. "The choices a modeler makes and the uncertainties that come with those choices convey how the modeler view the system and what they deem important, so it's crucial that students are able to articulate how their choices of data and methods reflect their values, and what the uncertainty tradeoffs are from simply choosing one set of modeling parameters and not another. In this context, we distinguish between intrinsic uncertainty—like future technology costs or human behavior—versus resolvable uncertainty, where you can improve results with better data or more advanced techniques. Students learn to look for both, describe them, and quantify the resolvable kind."

That distinction is more than academic. A professional who can identify what a model cannot know—and separate that from what it simply does not yet know—is far better equipped to communicate its limits to the people who will act on its conclusions.

From Data to Communication

To give students exposure to the kinds of data they will encounter in the field, the course draws on publicly available sources: EIA power plant data, the American Community Survey, Bureau of Labor Statistics labor data, and Miami's building energy benchmarking data, among others. Coding demonstrations walk students through complete analytical scripts, showing them how a real analysis is structured from start to finish.

"We aren't trying to train coders," Cong says. "We want to teach the logic of coding a modeling problem—how to think through it systematically, including deciding whether we need to manipulate the structure of a dataset for easier analysis, choosing appropriate visualizations to present the data and results, and reflecting on the modeling choices given in the coding demos."

Knowing how to build a model, however, is only part of what the energy workforce needs. Knowing how to explain it—to a city council member, a community association, a CFO, even a fellow modeler—matters just as much. Cong builds that skill deliberately. Students come from backgrounds in finance, humanities, engineering, and law, and she asks them to practice translating technical findings for audiences who may have none of those backgrounds.

"Communicating to stakeholders is difficult to practice," she says. "So I present a scenario with a model form and synthetic results, and have groups identify pieces of information from the model results that are the most important and relevant to their assigned stakeholder audience, also deciding and justifying what level of detail to present." By the end of the term, the improvement in fluency is unmistakable.

Ten Weeks of Progress

inline-met-class-sitting-600x400.jpg

MET courses are held in the Irving Institute's Call to Lead Lab. (Photo by Beam Lertbunnaphongs '25)

Ten weeks is a short time to go from an introduction to energy modeling to building an original analytical project from scratch. And yet, that is what students in this course did.

"The progress in ten weeks is astonishing," Cong says. "From new encounters with simple prompts to building their own models from scratch—and they all have unique spins. One group modeled the effectiveness of affordable housing decarbonization policies in two major cities, while another produced a physics-based power flow model to understand the impact of different configurations of new large loads." The range of approaches within a single cohort reflects both the diversity of student backgrounds and goals, and the flexibility of the course.

To push that learning further, Cong pairs project groups together before final presentations, requiring students to engage seriously with work outside their own chosen project topic. The conversations that follow—between a policy-focused team and a technically oriented one, for instance—surface insights that neither group would have reached alone, which ultimately improved each group's thinking about their own project. "If not for these exercises, they would not have had a chance to learn about each other's work beyond the one opportunity to watch the final presentations," she says. In a field where energy professionals routinely collaborate across specialties, that experience has practical value well beyond the classroom.

About Shuchen Cong

Shuchen Cong is a Postdoctoral Research Fellow at Dartmouth's Irving Institute for Energy and Society and a Lecturer at the Guarini School of Graduate and Advanced Studies. She holds a PhD and MS in Engineering and Public Policy from Carnegie Mellon University, where she also co-founded Peoples Energy Analytics, a university spinout that developed predictive tools to help utilities reduce customer default risk while supporting equitable access to energy.