A statistical model for risk management of electric outage forecasts

Li, H.; Treinish, Lloyd A.; Hosking, J. R. M.

doi:10.1147/jrd.2010.2044836

Cited by 38 publications

(37 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Their models are based on temperature, wind speed, and lightning. Li et al (2010) developed a Poisson regression model in a Bayesian hierarchical framework for predicting power outages with up to a 3-day lead time based on forecasts of severe weather events (e.g., hurricanes, tornados, and thunderstorms). DeGaetano et al (2008) developed an approach for forecasting ice accretion on electric distribution lines using the Weather Research and Forecasting model and an ice accretion model.…”

Section: Overview Of Hopmmentioning

confidence: 99%

“…Most previous methods for predicting power outages due to hurricanes (e.g., Liu et al 2005Liu et al , 2007Liu et al , 2008Han et al 2009a,b;Guikema et al 2010;Winkler et al 2010) have not adequately modeled the uncertainty in the hurricane forecasts or incorporated it into the outage forecasts. However, Li et al (2010) focused on predicting damage to the power system from all weather events. They provided confidence intervals on their damage predictions as well as worst-case scenarios.…”

Section: Incorporating Hurricane Fore-cast Error Information Into Thementioning

confidence: 99%

“…Li et al (2010) is the only prior study that has explicitly accounted for weather forecast uncertainty (i.e., differences between forecasted and observed weather variables). Although this study has focused on the needs of electric utilities, the approach and methods used for quantifying and representing uncertainty can be applied to model how other weather-sensitive enterprises such as agriculture, transportation, emergency management, insurance, and water resources will be effected by adverse weather events such as hurricanes (Li et al 2010). However, Li et al (2010) focused on predicting damage to the power system from all weather events.…”

Section: Incorporating Hurricane Fore-cast Error Information Into Thementioning

confidence: 99%

See 2 more Smart Citations

Incorporating Hurricane Forecast Uncertainty into a Decision-Support Application for Power Outage Modeling

Quiring¹,

Schumacher²,

Guikema³

2014

Bull. Amer. Meteor. Soc.

View full text Add to dashboard Cite

Uncertainty in power outage prediction models and other decision-support systems that are run prior to hurricane landfall can be improved using ensembles from the Monte Carlo wind speed probability model.T he National Hurricane Center (NHC) is responsible for issuing official track and intensity forecasts for all tropical cyclones' (TC) in the Atlantic and northeast Pacific basins. Official tropical cyclone forecasts have a variety of users from the public and private sectors and are often used to drive nonmeteorological models. These downstream models, called decision-support systems, provide useful information to stakeholders regarding tropical cyclonerelated impacts.Despite decades of research and model development that have led to a steady reduction in tropical cyclone track forecast errors and, recently, a more modest reduction in intensity forecast errors (Fig. 1), tropical cyclone forecasts aré not perfect. Although continued efforts will lead to better and better forecasts, forecast errors will always exist. This poses a particular challenge to those who use decision-support systems that depend on official TC forecasts, especially when it is unknown how forecast uncertainty combines with uncertainty of the models that are used in decision-support systems. For this reason, it is important not only to have ways of estimating and communicating uncertainly in TC track and intensity forecasts but also to understand how these errors affect decision-support system outputs.In this paper, we look at the impacts TC track and forecast errors have on a hurricane outage prediction model (HOPM). Starting at 24 h prior to landfall, a set of plausible tracks and intensity forecast scenarios based on the official NHC forecast are generated using the • ' In this paper, the term tropical cyclone is used to represent tropical storms, subtropical storms, and hurricanes.

show abstract

Section: Overview Of Hopmmentioning

confidence: 99%

Section: Incorporating Hurricane Fore-cast Error Information Into Thementioning

confidence: 99%

Section: Incorporating Hurricane Fore-cast Error Information Into Thementioning

confidence: 99%

See 1 more Smart Citation

Incorporating Hurricane Forecast Uncertainty into a Decision-Support Application for Power Outage Modeling

Quiring¹,

Schumacher²,

Guikema³

2014

Bull. Amer. Meteor. Soc.

View full text Add to dashboard Cite

show abstract

“…In addition, modeling utility-related problems is complex due to different interactions involved (e.g., tree conditions, soil saturation, infrastructure age). To address this complexity, an assortment of methods has been used for utility-related problems, including generalized linear models (GLMs; Li et al 2010), spatial and non-spatial generalized linear mixed models (GLMMs; Liu et al 2008), generalized additive models (GAMs; Han et al 2009a), classification and regression trees (CART; Quiring et al 2011), random forest ) and Bayesian additive regression trees (BART; Nateghi et al 2011). In addition to count data models, probabilistic models have also been coupled with physical models of the electric system with the aim to predict failures on both transmission and distribution lines (Mensah and Duenas-Osorio 2014).…”

Section: Introductionmentioning

confidence: 99%

Storm outage modeling for an electric distribution network in Northeastern USA

et al. 2015

View full text Add to dashboard Cite

The interaction of severe weather, overhead lines and surrounding trees is the leading cause of outages to electric distribution networks in forested areas. In this paper, we show how utility-specific infrastructure and land cover data, aggregated around overhead lines, can improve outage predictions for Eversource Energy (formerly Connecticut Light and Power), the largest electric utility in Connecticut. Eighty-nine storms from different seasons (cold weather, warm weather, transition months) in the period 2005-2014, representing varying types (thunderstorms, blizzards, nor'easters, hurricanes) and outage severity, were simulated using the Weather Research and Forecasting (WRF) atmospheric model. WRF simulations were joined with utility outage data to calibrate four types of models: a decision tree (DT), random forest (RF), boosted gradient tree (BT) and an ensemble (ENS) decision tree regression that combined predictions from DT, RF and BT. The study shows that the ENS model forced with weather, infrastructure and land cover data was superior to the other models we evaluated, especially in terms of predicting the spatial distribution of outages. This framework could be used for predicting outages to other types of critical infrastructure networks with benefits for emergency-preparedness functions in terms of equipment staging and resource allocation.

show abstract

“…These methods have limitation such as evolving power system inventory with time and presence of huge matrix of spatial correlation makes it computationally challenging. Poisson regression and Bayesian hierarchical network for risk management of power outages caused by extreme weather conditions is investigated in [6]. In this study, surface wind speed, gust speed, gust frequency, daily rainfall, daily minimum pressure and daily maximum and minimum temperature have been considered, while other weather factors such as lightning are excluded.…”

Section: Introductionmentioning

confidence: 99%

AdaBoost$^{+}$: An Ensemble Learning Approach for Estimating Weather-Related Outages in Distribution Systems

Kankanala

Das

Pahwa

2014

IEEE Trans. Power Syst.

View full text Add to dashboard Cite

Abstract-Environmental factors, such as weather, trees and animals are major causes of power outages in electric utility distribution systems. Of these factors, wind and lightning have the most significant impacts. The objective of this paper is to investigate models to estimate wind and lighting related outages. Such estimation models hold the potential for lowering operational costs and reducing customer downtime. This paper proposes an ensemble learning approach based on a boosting algorithm, ADABOOST + , for estimation of weather caused power outages. Effectiveness of the model is evaluated using actual data, which comprised of weather data and recorded outages for four cities of different sizes in Kansas. The proposed ensemble model is compared with previously presented regression, neural network, and mixture of experts models. The results clearly show that ADABOOST + estimates outages with greater accuracy than the other models for all four data sets.Index Terms-Artificial intelligence, ensemble learning, environmental factors, power distribution systems, power system reliability.

show abstract

A statistical model for risk management of electric outage forecasts

Cited by 38 publications

References 10 publications

Incorporating Hurricane Forecast Uncertainty into a Decision-Support Application for Power Outage Modeling

Incorporating Hurricane Forecast Uncertainty into a Decision-Support Application for Power Outage Modeling

Storm outage modeling for an electric distribution network in Northeastern USA

AdaBoost$^{+}$: An Ensemble Learning Approach for Estimating Weather-Related Outages in Distribution Systems

Contact Info

Product

Resources

About