IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 29, NO.2, MARCH 2014 889 Risk Analysis for Distribution Systems in the Northeast U.S. Under Wind Storms Gengfeng Li, Peng Zhang, Senior Member, IEEE, Peter B. Luh, Fellow, IEEE, Wenyuan Li, Fellow, IEEE, Zhaohong Bie, Senior Member, IEEE, Camilo Serna, and Zhibing Zhao Abstract-With the growing trend of extreme weather events in the Northeast U.S., a region of dense vegetation, evaluating hazard effects of wind storms on power distribution systems becomes increasingly important for disaster preparedness and fast responses in utilities. In this paper, probabilistic wind storm models for the study region have been built by mining 160-year storm events recorded in the National Oceanic and Atmospheric Administration's Atlantic basin hurricane database (HURDAT). Further, wind storms are classified into six categories according to NOAA criteria and IEEE standard to facilitate the evaluation of distribution system responses under different levels of hazards. The impacts of wind storms in all categories are accurately eval- uated through a Sequential Monte Carlo method enhanced by a temporal wind storm sampling strategy. Extensive studies for the selected typical distribution system indicate that our models and methods effectively reveal the hazardous effects of wind storms in the study region, leading to useful insights towards building better system hardening schemes. Index Terms-Critical facilities, distribution reliability, hard- ening planning, hazard, hurricane, wind storm. 1. INTRODUCTION F REQUENT wind storms have severely affected the North- east U.S. in the past few years. For instance, tropical storm Irene hit the State of Connecticut (CT) on August 28, 2011, causing sustained interruptions of electric service up to 11 days for over 800 000 customers and a total damage of about $200 million in CT [1], [2]. On October 22, 2012, hurricane Sandy swept the Northeast U.S. causing at least $50 billion in dam- ages to this area [3], [4], More than 850000 customers in all Manuscript received April 22, 2013; revised September 01, 2013 and October 09, 2013; accepted October II, 2013. Date of publication November 04, 2013; date of current version February 14, 2014. This work was supported in part by Northeast Utilities, USA. Some prepara- tory work was supported in part by National Natural Science Foundation of China under Grant 51077108. Paper no. TPWRS-00459-2013. G. Li is with State Key Laboratory of Electrical Insulation and Power Equipment, School of Electrical Engineering, Xi'an Jiaotong University, Xi'an, Shaanxi, 710049, China. He is also with the Department of Electrical and Computer Engineering, University of Connecticut, Storrs, CT 06269 USA. P. Zhang, P. B. Luh, and Z. Zhao are with the Department of Electrical and Computer Engineering, University of Connecticut, Storrs, CT 06269 USA (e-mail: peng@engr.uconn.edu). W. Li is with BC Hydro and Power Authority, Burnaby, BC V3N 4X8, Canada. Z. Bie is with State Key Laboratory of Electrical Insulation and Power Equipment, School of Electrical Engineering, Xi'an Jiaotong University, Xi'an, Shaanxi, 710049, China. C. Serna is with Northeast Utilities, Hartford, CT 06103-2818 USA. Color versions of one or more ofthe figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier IO.l109ITPWRS.2013.2286171 149 cities and towns served by Connecticut Light & Power suf- fered prolonged outages. The power outages lasted for over a month in some area of New York City. Analyzing power distri- bution system risks under extreme weather, therefore, is of sig- nificance in identifying system weaknesses, designing system hardening schemes and thus enhancing disasters preparedness in the Northeast. Impacts of extreme weather on power systems have previ- ously been studied. In IEEE Standard 346, weather conditions are divided into three categories: normal, adverse, and major storm disaster [5]. The National Oceanic and Atmospheric Ad- ministration (NOAA) developed the Saffir-Simpson Hurricane Wind Scale (SSHWS) that classifies hurricanes into 5 levels based on hurricane's sustained wind speed [6]. A three-state weather model was presented in [7] to incorporate failures occurred under major adverse weather conditions, following an observation that reliability evaluation results obtained without considering weathers could be optimistic and misleading. Ref- erence [8] presented a probabilistic hurricane simulation model established for assessing the Florida utility damage and risks under hurricanes. References [9] and [10] studied seasonal effects of wind and lighting on distribution system reliability, where time-varying failure rates based on partitioned weather severity levels were presented. Actually, the effects of extreme weather on distribution sys- tems are closely correlated to the region affected because of the specific elevation, terrain and vegetation in the particular region. The U.S. Northeast is a region with an appreciably high vege- tation coverage rate. For instance, the forest cover rate of CT even reaches to 75% [11]. Thus power outages in the Northeast regional distribution systems are largely caused by blow-over or failures of trees and poles during wind storms. However, an ap- propriate model for this region that reveals the impacts of wind storms has not been established. Moreover, the wind speeds of storms impacting the Northeast area fall into particular ranges, for which existing wind storm classifications in the literatures may be unsuitable. Main contributions of this paper include: 1) Probabilistic wind storm models including occurrence, intensity, and dura- tion models are established by mining the HURDAT database from NOAA. These models capture the effects of wind storms on the Northeast region, and are used to generate accurate wind storm samples for system risk assessment. 2) Outage event records from Northeast Utilities are used to parame- terize the weather-dependent component failure models. 3) An enhanced sequential Monte Carlo approach is developed to quantify system risks under six different categories based on a new storm classification criteria established for the Northeast 0885-8950 © 2013 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.orgipublications_standards/publicationsirights/index.html for more information.