Statistical & Optimization Approaches to Evaluating Energy Consumption, Supply, and Production Systems in a Decarbonized Future

<p dir="ltr">Decarbonization is a multi-faceted challenge. Deeply exploring multiple sectors of the energy system using statistical and optimization techniques within a larger interdisciplinary policy framework can provide rich insight into the opportunities facing consumers, planners, decision makers, and society at large as we pursue greater levels of carbon abatement. This thesis provides insights and recommendations to a series of technical challenges related to expanding renewable infrastructure deployment in a manner that considers the well-being of all in the process. </p><p dir="ltr">In the second chapter, I leverage utility smart-meter data in a differences-in-difference frame work to quantify the effect of LIHEAP funding on energy consumption behavior for 1,351 house holds in the Mid-Atlantic region of the U.S. Results reveal that LIHEAP funding leads to outcomes statistically indistinguishable from zero, although on average, we observe increases in energy consumption for homes that are electrically heated (37 ����ℎ ��������ℎ) and natural gas heated (0.30 ���������� ��������ℎ ). Additionally, wide confidence intervals in most of our findings signal substantial variability in household responses to LIHEAP. Lastly, under more granular analytic conditions, outcomes show that the effect of LIHEAP funding may result in a restriction of energy consumption. Based on results, we offer two policy recommendations: 1) enhance data sharing among utilities, researchers, and LIHEAP program administrators to facilitate mechanism-specific quantitative analyses and enable cross-regional comparisons of program impact; and 2) expand LIHEAP’s performance measures to include measures of energy consumption behavior. All of these outcomes prove important when considering the implications of residential electrification and its potential impacts v on low-income households. </p><p dir="ltr">In the third chapter, I shift into domestic transportation decarbonization, a key challenge facing the aviation sector in its pursuit of net-zero ����2 emissions by 2050. Using a mixed in teger linear optimization model in a multi-criteria decision analysis framework, I explored the cost and emissions tradeoffs of an optimally designed hydrogen supply chain for aviation in 2050 for the top 30 airports in the contiguous United States. Results reveal that optimally designed HSCs across all of our nominal analytic setups can be both cost and emissions competitive when compared to Jet-A and SAF incumbents, at 1 $/������−��2- 2 $/������−��2 . Through our model we also un cover that the footprint of most ��2 hub’s production extends well beyond that of the region of its origin, with the average pipeline transportation segment ranging from between 207 miles- 872 miles. Finally, we explore which network setups produce the absolute least cost, least emissions, or cost-resilient outcomes in an unknown future. I find that an HSC built under the assumptions associated with the curtailment setup in a future in which that reality persists in addition to fed eral production rebates results in the least cost solution. Regarding the least emissions option, I f ind that an HSC built under the assumptions associated with the rebate setup in a future with a fully decarbonized grid leads to the least emissions solution. As relates to cost resilience, we find that no additional information is needed from the present to design an HSC that can be within 27.7% of cost and 144.2% of emissions outcomes of the respective least cost and least emissions configurations. Through this work, I offer four key policy recommendations: 1) Continued grid decarbonization, which includes the expansion of transmission and generation, will play a key role in achieving some of the most aggressive decarbonization pathways for optimal HSCs; 2) Increased regional production flexibility, especially the expansion of biomass-derived hydrogen production, can help ameliorate regional production disparities and create a lower-carbon and cost-competitive pathway to hydrogen production that is not reliant on natural gas and carbon sequestration pathways; 3) creative funding structures, including power purchasing agreements or the expansion of curtailed energy at EV-truck charging stations, can go a long way in address vi ing the challenge of transportation operational expenses; 4) Cooperation across decarbonizing sectors like steel can leverage economies of scale to reduce annual HSC premiums and drive the sort of stable demand that is needed for robust hydrogen deployment. </p><p dir="ltr">Finally, in the fourth chapter of this dissertation, I focus on renewable electricity generation and reliability. Leveraging real-world load and solar availability data within a mixed integer linear programming model for CSP + GeoTES plants, this study uncovers a number of operational features of these renewable energy systems. At the daily level, we find that CSP + GeoTES plants are more sensitive to days with low solar availability when they have low recovery efficiency, resulting in operational outcomes that can appear intermittent. At a seasonal level, I observe that, on average, the GeoTES state of charge (SOC)ranges from hundreds of GWh of thermal energy in summer and fall months. However, the SOC decreases during the winter to tens of hours of GWh thermal due to energy being discharged from the GeoTES to compensate for low seasonal solar availability. When assessing the annual reliability performance of CSP + GeoTES systems across a suite of key system configurations, I find that a plant size of at least SM 3 and an maximum charge energy of at least 463 MW-t is the most robust system configuration that can meet a capacity factor threshold of at least 90% across different system reliabilities and annual solar availability levels. Ultimately, through this work we recommend that continued and increased investments be made into hybrid solar-geothermal solutions like CSP + GeoTES, as they represent a cost and time-to-deploy competitive solution to always on energy delivery.</p>