On the measurement of power grid robustness under load uncertainties

Power grids are unarguably the backbone of national infrastructures, providing vital support to economic growth. On the other hand, economic growth and the integration of plug-in hybrid electric vehicles have led to relentless power demand increase. Unfortunately, the update of power grids cannot always keep pace with power demand growth, which imposes potential threats on the reliable operation of power grids. When this occurs, the grids may still operate well under normal conditions, but when some loads change suddenly, they may fail much more easily. It is, therefore, imperative to establish a quantitative method to evaluate the robustness of power grids under load uncertainties, which can benefit grid planning, upgrade, and optimization. Unfortunately, despite the importance of this problem, it remains open in the literature. In this paper, we formally formulate the problem of power grid robustness and propose an efficient framework to measure it based on two sampling techniques: StopSign and Interval. The performance of the proposed framework has been evaluated on IEEE standard bus systems. To the best of the authors' knowledge, this is the first in-depth work on power grid robustness estimation under load uncertainties.

I. Introduction

Power grids are critical infrastructures to deliver electricity for consumers and provide a vital support to economic growth. On the other hand, economic growth and the integration of plug-in hybrid electric vehicles result in a rapid increase in power demand. According to the data released by the North American Electric Reliability Corporation (NERC) in 2015 [1], the electricity demand has increased by more than 70% over 15 years, as illustrated in Figure 1. Unfortunately, upgrading power grids (e.g., adding new transmission lines or upgrading the existing ones) cannot always keep pace with the power demand growth, partially because of the high cost associated therewith. For example, the American Recovery and Reinvestment Act of 2009 allocated $4.5 billion to the U.S. Department of Energy (DOE) for investments in upgrading infrastructures of power grids [2]. Consequently, the grids may still operate well under normal conditions, but when one or more loads swing beyond their normal ranges (which can occur due to unforeseen activities), they may fail much more easily. Investigating the robustness of a given power grid structure under load uncertainties are thus highly desirable.