Diagnosing Thunderstorms Induced Power Outages with the Rapid Refresh Model

Search for this publication on Google Scholar
Cucinotta, R. (2019). Diagnosing Thunderstorms Induced Power Outages with the Rapid Refresh Model. Unc Charlotte Electronic Theses And Dissertations.
Download PDF

Analytics

84 views ◎
62 downloads ⇓


Abstract

Duke Energy customers frequently experience electrical power outages caused by thunderstorms that produce strong surface winds and subsequently damage power distribution infrastructure. To better anticipate such outages, forecasters would benefit from outage-focused guidance regarding storm strength and timing so proper outage-mitigation can be initiated prior to each event. This study identified meteorological parameters that could best predict total daytime and nighttime power outages across each Duke Energy service area using generalized linear models (GLMs). A total of 392 event dates were stratified with regard to five service areas, two seasons, six convective modes, and three dominant severe weather types. Daytime and nighttime GLMs were attempted for each stratification with more than 10 event dates and the resulting GLMs were deemed "operational" if predictors remained statistically significant (p-values < 0.05) for ten trials using a randomly-selected portion of the training dataset. In total, 58 operational GLMs were developed using 24 unique parameters as predictors. The most common predictors were low-level vertical velocity and composite parameter used to predict large hail. The resulting operational GLMs can be implemented within an ensemble framework to provide Duke Energy with a total outage estimate for each service area.

Details
Author

Cucinotta, Rachel

Title

Diagnosing Thunderstorms Induced Power Outages with the Rapid Refresh Model

Physical Description

1 online resource (193 pages) : PDF

Date

2019

Degree Granting Institution

University of North Carolina at Charlotte

Abstract

Duke Energy customers frequently experience electrical power outages caused by thunderstorms that produce strong surface winds and subsequently damage power distribution infrastructure. To better anticipate such outages, forecasters would benefit from outage-focused guidance regarding storm strength and timing so proper outage-mitigation can be initiated prior to each event. This study identified meteorological parameters that could best predict total daytime and nighttime power outages across each Duke Energy service area using generalized linear models (GLMs). A total of 392 event dates were stratified with regard to five service areas, two seasons, six convective modes, and three dominant severe weather types. Daytime and nighttime GLMs were attempted for each stratification with more than 10 event dates and the resulting GLMs were deemed "operational" if predictors remained statistically significant (p-values < 0.05) for ten trials using a randomly-selected portion of the training dataset. In total, 58 operational GLMs were developed using 24 unique parameters as predictors. The most common predictors were low-level vertical velocity and composite parameter used to predict large hail. The resulting operational GLMs can be implemented within an ensemble framework to provide Duke Energy with a total outage estimate for each service area.

Genre

masters theses

Subjects--Topics

Meteorology
Atmospheric science

Degree

M.S.

Keywords

Generalized Linear Regression
Microbursts
Power Outages
Warm Season Convection

Subject Area

Earth Sciences

Advisor(s)

Eastin, Matthew

Committee Members

Lanham, Stanton
Shirley, Terry
Magi, Brian

Degree Note

Thesis (M.S.)--University of North Carolina at Charlotte, 2019.

Rights Statement

This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s). For additional information, see http://rightsstatements.org/page/InC/1.0/.

Rights Holder Information

Copyright is held by the author unless otherwise indicated.

Identifier

Cucinotta_uncc_0694N_12208

Permalink

http://hdl.handle.net/20.500.13093/etd:2616