Redefining Contingency Analysis for Development of Resilient Cyber Physical Power System
By Desh Deepak Sharma
Resiliency has been a critical issue in the power grid in order to avoid cascading failure, improve ability to recover quickly after high impact disturbances and mitigate such type of events. In a power system, the equipment/transmission line loading must be within its specified operational limit capacity. The contingencies which may violate the specified limits must be identified. The limiting values of network operation variables such as voltage and frequency must be identified in order to avoid unacceptable loading of the equipment and loss in the useful life of the equipment. The N-1, N-2 contingencies must be predicted in real time before such situations might occur.
With the adoption of information and communication technology (ICT) in power grid, a cyber physical of power system (CPPS) is developed. In CPPS, there is possibility of machine-to-machine communication in order to avoid any outage which may lead to cascading failures. The complex algorithms are to be run on digital computers which may reside far away from the location of the fault. In the presence of such type of architectures, new distributed algorithms must be developed for a resilient cyber physical power system.
Over loading is a one of major factor of cascading failures in the power system. The voltage and current stresses of the elements may cause the sequence of failures. These stresses can be calculated for analysis by using the deterministic circuit equations. There is big challenge to determine the time duration between failures. The research efforts are expected in development of mathematical modelling of these time durations in cyber physical of power system in order to avoid cascade failures and improve the resiliency of power system. The network based stochastic models for time durations between failures are suggested to understand the dynamics of cascading failures.
The CPPS consists of power apparatus as physical part, and cyber part which does state monitoring, communications and control of physical network. In this CPPS, the interdependency of these two kinds of networks must be analyzed in terms of cascade failures. For example, effects due to failure of power equipment and, subsequently, dispatching of the respective data are to be assessed, simultaneously. The dynamics of cascading failure can be considered as a sequence of state transitions of the physical and cyber nodes. The cyber layer supports the power system operations via the communication network especially during occurrence of faults in power grid. As the power network and dispatch data network are interdependent, the failure of one network might affect another network. Thus, the modeling of this interdependency based on dynamic power flow is required to be done in order to improve security and reliability of the system. The one-to-one interdependent network assumes that both the power network and cyber network comprises equal number of nodes. The studies based on percolation theory have been done for the analysis of cascade failures. According to these studies, the cascading failure transition is the first order phenomena in one-to-one interdependent network while this is the second order phenomena in an isolated network. In the multi-interdependent network, the interdependency between power grid and communication network is one to multiple and the coupling between these two networks is more complicated. In such systems, the research works are needed to study the resiliency against failures along with redundancy in the system. Furthermore, a model has to be developed for the relationship between resiliency and the redundancy of the system. A real- time event of cascading failure is to be created on an experimental set up. Based on the experiments, this is shown that there is possibility of cascade failure on occurrence of sharp increment in the angular separation between generators. The dispatching data network is of two types as double star and mesh. The good dispatching data network improves the monitor and control part of the whole system. But, the information may be lost and power appliance may be affected due to a cyber-attack. The topological characteristics of power grid with consideration of power grid as complex graph are designed for the analysis of cascade failure. The characteristics as average degree, cluster coefficient, average path length and algebraic connectivity are defined to analyze the outage phenomena in power grid.
Edited by Mehmet Cintuglu
Desh Deepak Sharma is an Associate Professor with the Department of Electrical Engineering, M. J. P. Rohilkhand University, Bareilly, India. He did M.Tech from NIT Kurukshetra, PhD from IIT Kanpur. His research interests include demand side management, application of data mining techniques in load profiling and distribution system, distributed control schemes for distributed generation and energy storage systems, multi-agent system for power system, smart grid as cyber-physical system, and cyber security issues in power system.