Conference Paper

A probabilistic loading-dependent model of cascade failure and possible implications for blackouts

Dept. of ECE, Wisconsin Univ., Madison, WI, USA
DOI: 10.1109/HICSS.2003.1173909 Conference: System Sciences, 2003. Proceedings of the 36th Annual Hawaii International Conference on
Source: IEEE Xplore

ABSTRACT Catastrophic disruptions of large, interconnected infrastructure systems are often due to cascading failure. For example, large blackouts of electric power systems are typically caused by cascading failure of heavily loaded system components. We introduce the CASCADE model of cascading failure of a system with many identical components randomly loaded. An initial disturbance causes some components to fail by exceeding their loading limit. Failure of a component causes a fixed load increase for other components. As components fail, the system becomes more loaded and cascading failure of further components becomes likely. The probability distribution of the number of failed components is an extended quasibinomial distribution. Explicit formulas for the extended quasibinomial distribution are derived using a recursion. The CASCADE model in a restricted parameter range gives a new model yielding the quasibinomial distribution. Some qualitative behaviors of the extended quasibinomial distribution are illustrated, including regimes with power tails, exponential tails, and significant probabilities of total system failure.

Download full-text


Available from: Ian Dobson, Sep 26, 2015
27 Reads
  • Source
    • "Cascading failures propagating in power grids may lead to catastrophic blackouts, e.g. the blackouts in North America on 14th August 2003 [2], in Brazil on 10th November 2009 [3], and in India on 30th and 31st July 2012 [4]. To analyze cascading failures and evaluate their influences on power grids, several models have been proposed, such as the Manchester model [5], CASCADE model [6], hidden failure model [7]-[9], branching process model [10], [11], and OPA 1 model [12]-[16]. Manchester model considers overloads, relay malfunction, and transient instability, corresponding to a relatively detailed simulation. "
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper develops a test bed based on the Northeastern Power Coordinating Council (NPCC) 48-machine, 140-bus power system model for simulating cascading failures. Then, the paper uses the test bed to demonstrate the interaction model recently proposed by [1] for analyzing cascading failures, identifying key linkages between component failures, and predict propagation of failures. For the NPCC test bed, a new efficient method is proposed to determine line flow limits as critical parameters in cascading failures simulation. Then, those limits are utilized in AC-OPA to generate a database of cascades. Finally, the interaction model is derived based on the database to extract key knowledge on cascades, which can be used for online analysis and simulation of cascading failures.
    IEEE Power and Energy Society General Meeting, Denver CO; 07/2015
  • Source
    • "Several models have been proposed, such as CASCADE model [1], Branching process model [2], Hidden failure model [3]-[5], OPA model [6], improved OPA model [7], and Manchester model [8]. However, these models do not consider the slow process at the beginning of blackout. "
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, a blackout model that considers the slow process at the beginning of blackouts is proposed based on the improved OPA model. The model contains two layers of iteration. The inner iteration, which describes the fast dynamics of the system, simulates the power system cascading failure, including the tree contact and failure of lines caused by heating. Compared with the improved OPA model, the outer iteration, which describes the long-term slow dynamics of the system, adds the simulation of tree growth and utility vegetation management (UVM). Moreover, the proposed model also improves the simulation of protective relays and the dispatching center and makes them closer to practical conditions. The effectiveness of the proposed model is verified by the simulation results of Northeast Power Grid of China.
    Power and Energy Society General Meeting, 2012 IEEE; 07/2012
  • Source
    • "Widespread blackouts can occur due to the cascading failure of components. Typically, an initial disturbance causes a component to fail, which results in the overloading of other components, thus, cascading failure of further components becomes likely [33]. The ability to quantify cascading blackout risk utilizing the probabilistic technique offers ways to monitor power system reliability and thus to minimize as much as possible these catastrophic failures. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Power system security is defined as the ability of a power system to maintain supply without unduly allowing network variables to stray from prescribed ranges. Traditionally, security has been assessed using deterministic criteria e.g. 'N-1' or 'N-2' under prescribed severe system loading levels. However, such worst-case deterministic approach does not provide explicitly an assessment of the probability of failure of the system, and the likelihood of the outages is treated equally. This approach may result in either over or under estimation of system planning reinforcement requirements and, hence, a corresponding excessive or insufficient system security. On the other hand, probabilistic security assessment may offer advantages by considering (i) a statistical description of the performance of the system over an annual cycle together with (ii) the application of historical fault statistics that provide a measure of the probability of faults leading to systems outages. This paper reviews approaches to probabilistic security assessment. Such approaches include (i) identifying various security indices, (ii) reviewing different techniques, (iv) utilizing Monte Carlo simulation, and (iv) integrating deterministic & probabilistic techniques. The application of probabilistic security assessment in real-time operation and system planning is further considered in detail. Preliminary application of the security assessment is presented to the Dubai Electricity and Water Authority (DEWA) transmission network.
    Universities Power Engineering Conference (UPEC), 2012 47th International; 01/2012
Show more