Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


Shield_Thesis_Final.pdf (5.34 MB)

ETD Abstract Container

Abstract Header

Predictive Modeling of Thunderstorm-Related Power Outages

Shield, Stephen, Shield
http://rave.ohiolink.edu/etdc/view?acc_num=osu152951430854521

Abstract Details

2018, Master of Arts, Ohio State University, Geography.
Each year in the United States weather related power outages result in billions of dollars of restoration costs and economic losses. Utility companies, emergency management agencies, and other public and private entities affected by power outages attempt to anticipate and mitigate the effect of these outages by utilizing power outage prediction models. These models are typically developed for either a combination of weather events or specialized for specific weather events like tropical cyclones. Despite the fact that thunderstorms account for almost half of major power outage events, development of specialized models for thunderstorms is at an early stage. In this research we use the random forest machine learning technique to develop specialized models for thunderstorm related power outage events. The models are trained using power outage data from 31 thunderstorm events along with 75 predictor variables that include forecast weather conditions and environmental variables that have been found to improve power outage prediction models in past research. This is done at three spatial scales (~3 km, 15 km, and 40 km) to account for the potential influence of spatial scale on model performance. Results showed spatial scale is an important factor in both the level of model performance as well as the importance of certain variables. At finer scales environmental variables had high importance and convective hazard probabilities issued by NOAA’s National Weather Service Storm Prediction Center (SPC) had low importance. As the spatial resolution becomes coarser the importance of variables reverses with environmental variables becoming less important and convective hazard probabilities becoming more important. Analysis of model performance showed that a two-stage random forest model at a 40 km spatial resolution had better performance than the models at finer spatial resolutions. However, it still had a tendency to under predict outages in the most intense thunderstorm events. Additionally, the lack of importance among environmental variables has negative implications for geographic scalability of the 40 km model.
Steven Quiring (Advisor)
Jay Hobgood (Committee Member)
Elisabeth Root (Committee Member)
86 p.
Geography
Predictive Modeling, Power Outages, Thunderstorms

Recommended Citations

Citations

  • Shield, Shield, S. (2018). Predictive Modeling of Thunderstorm-Related Power Outages [Master's thesis, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu152951430854521

    APA Style (7th edition)

  • Shield, Shield, Stephen. Predictive Modeling of Thunderstorm-Related Power Outages . 2018. Ohio State University, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu152951430854521.

    MLA Style (8th edition)

  • Shield, Shield, Stephen. "Predictive Modeling of Thunderstorm-Related Power Outages ." Master's thesis, Ohio State University, 2018. http://rave.ohiolink.edu/etdc/view?acc_num=osu152951430854521

    Chicago Manual of Style (17th edition)

osu152951430854521
633
© 2018, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.