A framework for Assessing the Reliability of Grid Networks by Modelling the Cyber-Physical Interdependencies of Dynamic Line Rating Components

Abstract

This study introduces a method for assessing the steady-state availability of a power grid network comprising dynamic line rating (DLR) and other cyber-physical elements. The network's reliability is modelled as a Markov process representing the availability of both the cyber network, including DLR sensors and the electrical power system component. The study determines the failure rate of the combined systems using a Markov chain transiting between four states. It also recommends a placement method for positioning DLR sensors in the cyber network, which involves utilizing a clustering algorithm and Monte Carlo simulation. The availability of the electrical power system network at normal operating conditions is evaluated from the loss of load probability and the repair rate of the network, while that of the cyber network is determined by the latency of communication and the cyber network repair rate. The steady-state availability of the combined cyber-physical power system at base load was found to be 0.5828 (5105 hours) per year. By systematically decreasing the failure rate of the cyber network to 75%, the availability of the power system network increased to 0.865 (7577 hours) per year.