Doctoral Dissertation: A Low Carbon Economy: Addressing Technical Challenges Spanning the Electricity and Steel Sectors
The Intergovernmental Panel on Climate Change states that in order to reduce the most extreme consequences of climate change we must reduce global greenhouse gas (GHG) emissions by 41-72% by 2050 from 2010 levels. For this to happen, all sectors must be included in the effort to decarbonize. Two major sources of anthropogenic emissions are the electricity and industrial sectors, combined accounting for 50% percent of U.S. GHG emission in 2017. This dissertation addresses challenges with decarbonization in the electricity sector and outlines potential decarbonization pathways in the U.S. steel sector, which accounts for 25% of industrial sector CO2 emissions.
This dissertation addresses two electricity sector challenges, which stem from the inherent variability and uncertainty of wind and solar generation, which we specify as variable renewable energy (VRE). The first is the need for sustainable frequency regulation resources, which increase with the addition of VRE. Energy storage, specifically lithium ion batteries, is an attractive option to supply additional frequency regulation. However, they may not reduce the systems environmental impacts. In order to quantify the range of potential environmental impacts Chapter 2 employs a life cycle assessment (LCA) framework, which couples cradle-to-gate and end-of-life LCA data on lithium ion batteries with a unit commitment and dispatch model. The model is run on a 9-bus power system with lithium ion batteries used for frequency regulation. Chapter 2 finds that the sustainability of lithium ion batteries in this application depends on the grid mix and that the intensity of use does not increase the upstream and end-of-life impacts beyond the use-phase.
The second challenge with the addition of VRE this dissertation addresses is the need to improve day-ahead generator scheduling to prevent unplanned startups and shutdowns (i.e., commitment error) caused by the increased variability in load. Chapter 3 uses a regression to quantify the relationship between net load shape, VRE generation, and commitment error with the goal of predicting commitment error so that day-ahead schedules can be improved and system inefficiency reduced. Chapter 3 finds statistically significant relationships between net load, ramp rate, portion VRE generation, and commitment error. Using these relationships, the regression models were able to explain 21-57% of the variability in commitment error.
This dissertation also analyzes options for U.S. steel sector decarbonization. Steel use spans almost all sectors and its use is correlated to economic growth. However, its production is carbon intensive. Through examination of available and developing production technologies in combination with steel flow modeling Chapter 4 outlines potential pathways the U.S. could take to cut emissions allocated to its steel consumption 70% by 2050 from 2010 levels. There are a number of actions the U.S. could take to reduce CO2 emissions, however, Chapter 4’s results indicate that, given available technologies, steel stocks per capita, in turn demand, must be reduced in the near term to have a chance of cutting CO2 emissions 70%.
The challenges in decarbonizing the electricity sector are focused on the supply side, while this dissertation recommends demand side changes for the steel sector. This is a result of the stark difference in availability of low carbon technologies. Although there remain outstanding challenges with electricity decarbonization they are technological or market based both being easier to address than the demand changes required in the steel sector, which will likely require top down regulation and incentives to achieve.
Chairs: Prof. Shelie Miller and Prof. Steve Skerlos