Severe weather, utility spending, and the long-term reliability of the U.S. power system

Abstract

There has been a limited amount of peer-reviewed literature on long-term trends in electricity reliability including the underlying factors that impact reliability across the United States. In this analysis, we considered up to 16 years of data from 203 U.S. utilities-representing about 70% of electricity sales. Annual frequency of interruptions for an average customer-at the regional and U.S. national-level-has generally decreased over this timeframe. But we do not find that there is a statistically significant trend in the annual duration of interruptions for an average customer. We find that more explicit measures of severe weather are correlated with reliability. We are able to explain 7% and 16% of past variation in the reliability metrics system average interruption duration and frequency indices, respectively, is due to severe weather-a significant improvement over earlier studies. We find that current year spending by utilities is correlated with worse reliability and that cumulative spending over the preceding three years is correlated with better reliability. Finally, we demonstrate that using a statistical instrument to represent the annual frequency of interruptions for an average customer can greatly improve analysis of trends in the annual duration of interruptions for an average customer. 1. Introduction Power interruptions are caused by a number of different factors including weather-related impacts, electrical equipment faults or failure, and, indirectly, spending strategies on power system infrastructure, operations, and maintenance. The U.S. Department of Energy (DOE) reports that extreme weather is the most commonly-reported cause of power interruptions with the frequency of these extreme events increasing significantly over the last two decades [1]. Power interruptions significantly impact economic activity. A recent study estimates that sustained power interruptions cost an average of $44 billion annually in the U.S. alone [2]. The 2017 Atlantic hurricane season is just one example of recent extreme weather that has caused long duration, widespread power interruptions. During this season, six hurricanes were classified as "major" (i.e., Saffir-Simpson Category three storms or higher). For perspective, the long-term average number of major hurricanes since 1851 is six per decade [3]. In August 2017, Hurricane Harvey flooded many parts of Texas and Louisiana where some locations received as much as five feet of rain, resulting in power interruptions that impacted more than 330,000 customers [4]. During this storm, substations were flooded, utility poles were toppled, and there was extensive damage to other critical energy and electricity infrastructure [5, 6]. The following month, Hurricane Irma caused widespread damage to the Florida Keys and resulted in power interruptions for nearly two million customers [7]. Less than a week later, Hurricane Maria-the second category five storm of the season-caused widespread devastation across the U.S. territory of Puerto Rico. This storm destroyed the is...

A Stochastic Planning Model for Improving Resilience of Distribution System Considering Master-Slave Distributed Generators and Network Reconfiguration

“…Block index e Energy storage index b, b ' Bus indices k Microgrid index l, l ' Distribution Extreme weather conditions in recent years have resulted in long and widespread electricity service interruptions with enormous economic losses [1]. For example, Hurricane Harvey resulted in power interruptions that affected several million customers in Texas for 14 days [2].…”

“…In this section, the objective function minimizes the investment cost in the first stage and the expectation of load shedding cost under line damage scenarios in the second stage, shown in (1). It is worth mentioning that the load shedding cost is used to assess the distribution system performance during extreme events, and this index has been widely utilized to evaluate the resilience of the distribution system [27,28].…”

Large-scale data analytics for resilient recovery services from power failures

“…9,10 The long interruption time is found to be particularly damaging to customers. 11,12 The policy itself, however, has been largely sidelined by the prior work. Further, large-scale power failures induced by weather disruptions happen with an increasing severity.…”

Reliability of a Power System With High Penetration of Renewables: A Scenario-Based Study

“…Reliability studies are generally performed in the planning or short-term scheduling stages. Power system reliability indicates the overall ability of the power system to deliver its power generation to end-users [2][3][4][5]. Typical power system reliability studies are categorized into two groupsthose that address adequacy and those that address security [6].…”

“…This additional but essential factor (spinning reserves) was generally ignored in previous studies. (2) The fuzzy-c-means algorithm is utilized to develop the Markov models of load, wind power and photovoltaic power in the individual state spaces, which are further transformed into an aggregated scenario space. (3) Rather than solving the LOLP problem in a chronology space, the proposed method deals with the LOLP in a scenario space, significantly reducing the computational time that is required by MCS.…”