Preprints
https://doi.org/10.5194/egusphere-2022-975
https://doi.org/10.5194/egusphere-2022-975
 
10 Oct 2022
10 Oct 2022
Status: this preprint is open for discussion.

Probabilistic and Machine Learning Methods for Uncertainty Quantification in Power Outage Prediction due to Extreme Events

Prateek Arora1 and Luis Ceferino1,2 Prateek Arora and Luis Ceferino
  • 1Civil and Urban Engineering Department, New York University, Brooklyn, NY, 11201, USA
  • 2Center for Urban Science and Progress, New York University, Brooklyn, NY, 11201, USA

Abstract. Strong hurricane winds damage power grids and cause cascading power failures. Statistical and machine learning models have been proposed to predict the extent of power disruptions due to hurricanes. Existing outage models use inputs including power system information, environmental, and demographic parameters. This paper reviews the existing power outage models, highlighting their strengths and limitations. Existing models were developed and validated with data on a few utility companies and regions, limiting the extent of their applicability across geographies and hurricane events. Instead, we train and validate these existing outage models using power outages for multiple regions and hurricanes, including Hurricanes Harvey (2017), Michael (2018), and Isaias (2020), in 1,833 cities along the U.S. coastline. The dataset includes outage data from 39 utility companies in Texas, 5 in Florida, 5 in New Jersey, and 11 in New York. We discuss the limited ability of state-of-the-art machine learning models to (1) make bounded outage predictions, (2) extrapolate predictions to high winds, and (3) account for physics-informed outage uncertainties at low and high winds. For example, we observe that existing models can predict outages as high as 25 times more than the number of customers and cannot capture well the outage variance for wind speeds over 70 m/s. Finally, we present a Beta regression outage modeling framework to address the shortcomings of existing power outage models.

Prateek Arora and Luis Ceferino

Status: open (extended)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on egusphere-2022-975', Anonymous Referee #1, 11 Nov 2022 reply

Prateek Arora and Luis Ceferino

Prateek Arora and Luis Ceferino

Viewed

Total article views: 397 (including HTML, PDF, and XML)
HTML PDF XML Total Supplement BibTeX EndNote
296 92 9 397 27 3 2
  • HTML: 296
  • PDF: 92
  • XML: 9
  • Total: 397
  • Supplement: 27
  • BibTeX: 3
  • EndNote: 2
Views and downloads (calculated since 10 Oct 2022)
Cumulative views and downloads (calculated since 10 Oct 2022)

Viewed (geographical distribution)

Total article views: 386 (including HTML, PDF, and XML) Thereof 386 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
1
 
 
 
 
Latest update: 06 Dec 2022
Download
Short summary
Power outage models can help the utilities in managing risks for outages from hurricanes. Our article reviews the existing outage models during hurricanes and highlights their strengths and limitations. Existing models can give erroneous estimates with outage predictions larger than the number of customers, struggle with predictions for catastrophic hurricanes, and do not represent the uncertainties of infrastructure failure well. We conceptualize a new model that overcomes these challenges.