Abstract
Introduction. This Undergraduate Thesis Portfolio presents two complementary investigations of the electrical infrastructure that sustains modern American life. The technical capstone develops a high-resolution digital twin framework for evaluating price-based demand response policies at the household and substation level, using Virginia’s Eastern Shore as a case study. The STS research paper examines the social consequences of Virginia’s rapid data center expansion, focusing on how the growth of artificial intelligence infrastructure distributes costs and reliability risks unevenly across the state’s communities. Although the two projects approach the grid from different angles—one engineering a computational tool for utility planners, the other analyzing the political and social forces shaping infrastructure decisions—they share a common subject: the Virginia electrical grid, which must simultaneously absorb hyperscale data center load growth and deliver affordable, reliable power to residential customers. Together, the projects illustrate how technical modeling and sociotechnical analysis can jointly inform more resilient and equitable energy policy during a period of unprecedented demand growth.
Capstone Project Summary. The U.S. power grid faces growing reliability challenges as rising peak demand, renewable intermittency, and baseload capacity retirements strain existing infrastructure. While long-term solutions require grid expansion, near-term resilience depends on optimizing current resources. Price-based demand response (PBDR) incentivizes households to shift electricity use away from peak periods, but existing evaluations aggregate load reductions across entire service territories and obscure impacts on specific grid assets and communities. To address this gap, the capstone develops a high-resolution digital twin framework that couples a synthetic population, hourly electricity demand profiles, and a distribution network to derive system-level reductions from individual households, together with a stochastic adoption model and demand reallocation algorithm that captures how household demographics drive load shifts across the network. A case study of Virginia’s Eastern Shore under two candidate PBDR policies produced peak load reductions of four to five percent, consistent with empirical literature, while resolving shifts at the household and substation levels. By linking demographic characteristics and load-shifting behavior to substation-level outcomes, the framework can inform more resilient demand-side energy policy and is extendable nationwide.
STS Research Paper Summary. The rapid expansion of data centers in Northern Virginia has created a situation in which digital infrastructure development is advancing faster than the electrical infrastructure required to support it. Evidence from energy research, infrastructure studies, and social science literature suggests that this mismatch is likely to produce several significant social consequences, including rising electricity prices, an elevated risk of grid failures, and an unequal distribution of these impacts across communities. The paper draws on two complementary frameworks from Science, Technology, and Society scholarship—Everett Rogers’ Diffusion of Innovation theory and the Social Construction of Technology perspective—to argue that the data center boom is not an inevitable consequence of technological progress but a socially constructed process whose pace and distributional effects reflect specific political and economic decisions. Using documentary research and comparative case analysis, the paper finds that the infrastructure strain, higher utility rates, and outage risks associated with data center load growth are most likely to fall on economically vulnerable communities that have the fewest resources to adapt.
Concluding Reflection. Working on both projects simultaneously reshaped how I understood each one. The capstone trained me to think about the grid as a quantifiable system of substations, loads, and optimization objectives, while the STS paper pushed me to ask who the model ultimately serves, whose demand it counts, and whose interests shape the policies it evaluates. These questions changed the capstone. Encountering energy-justice scholarship while writing about Virginia’s data center corridor made the decision to preserve household-level and substation-level resolution feel less like a modeling preference and more like a deliberate commitment to making distributional impacts visible to utility planners. Conversely, working through load-flow simulations made the abstract claims in the STS paper—that grid stress and cost recovery are unevenly distributed—feel tangible and mechanical rather than merely rhetorical. The combination also clarified my professional orientation: engineers build the models that quantify tradeoffs, but the tradeoffs themselves are moral and political. Producing a technical tool and a social analysis of the same infrastructure at the same time taught me that the two tasks are inseparable.
