Abstract
The United States electricity grid is one of the most crucial infrastructures in daily life, underpinning many domains, including transportation, health, communication, agriculture, and many others. Yet, the decisions that shape its development and governance are rarely visible or understood by the public. My technical capstone, Holding the Line: A Digital Twin Framework Supporting Policy Design for U.S. Power Grid Resilience, addresses how existing demand response evaluations look at aggregate load over a region. This is necessary because when load is analyzed at a general level, it masks the impacts on specific grid assets like substations. My STS paper, Defining Cost and Resilience: Stakeholder Tensions in U.S. Grid Modernization, investigates how competing stakeholder definitions shape grid governance decisions. This is motivated by the idea that grid failures are not simply technical breakdowns, but rather are shaped by social, political, and economic structures. Together, the papers explain the inner workings of how grid decisions are made and who is most impacted. The technical project builds a tool to make policy more precise, while the STS paper explains why governance frameworks often fail to protect the communities that are most vulnerable to grid disruptions.
The technical capstone project addresses the gap between system-level modeling and the outcomes relating to substations that matter for distribution planning. Our team constructed a high-resolution digital twin of Virginia’s Eastern Shore, combining a synthetic household population of 28k+ households with annual hourly electricity load profiles and a synthetic distribution network of 21 substations. An adoption model assigned each household a participation probability based on demographic factors in two policies. Monte Carlo simulation was then used to create probabilistic substation-level load estimates.
The results demonstrate that established demand response effects can be reproduced at household and substation resolution without loss of fidelity. The TOU scenario achieved a mean peak reduction of 3.6% across substations, and the Tempo scenario achieved 4.1% reduction on Red days, both consistent with empirical literature. Importantly, the framework revealed dynamics that would have otherwise been hidden in an aggregate model, confirming that substation-level outcomes may be obscured if planning is conducted only at the system-wide level and that effective policy requires the granularity that household-level resolution provides.
The STS paper asks how competing stakeholder definitions of resilience and cost-efficiency become institutionalized in U.S. grid governance and whose definitions get left out. This question matters because grid governance debates are about how value, cost, and protection are defined and considered in policy. Using the Social Construction of Technology (SCOT) framework, which says that technologies are shaped by the values and power of the groups that create and use them, the paper analyzes academic literature, FERC reports, Public Utility Commission decisions, and news coverage. Two case studies further this analysis, both the Winter Storm Uri in Texas (2021) and California's Public Safety Power Shutoff (PSPS) program.
This paper finds that dominant definitions of resilience and cost-efficiency reflect the priorities of utilities and regulators while marginalizing vulnerable consumers. In Texas, the failure to winterize power plants following prior recommendations caused over 200 deaths and financial losses of up to $130 billion. California's PSPS program similarly imposed severe burdens, particularly given that low-income households already spend 8.1% of their income on energy costs compared to 2.3% for non-low-income households. The paper concludes that improving grid governance requires expanding official definitions of resilience beyond engineering metrics to include community-level vulnerability.
