Resilience is the ability of any system to reduce the chances of shocks, absorb the shock if it occurs and recover quickly if a shock causes disruption. Additionally, the main features of resilience are redundancy, robustness, rapidity and resourcefulness. Power systems provide essential services in the modern societal setup, however, they are susceptible to various hazards, such as; floods, hurricanes, earthquakes and terrorist attack – among others. Moreover, power systems are increasingly becoming more complex and larger, thereby, making power disruptions more damaging due to the interdependency of systems. Therefore, power systems should be resilient. In a bid to quantify the resilience of a power system, several measures have been proposed, such as the resilience triangle paradigm and probabilistic model approaches. Unfortunately, most of the published work based on the aforementioned approaches focusses on the definitions, modelling and analysis of system resilience with minimal consideration of the system’s empirical evaluation and quantification based on obtained data. Thus, it is imperative that the gap existing between the methodological studies and the practical applications be closed.
To this effect, under the funding from the CREATE program on Future Resilience Systems, a team of researchers at National University of Singapore led by Professor Loon Ching Tang from the Department of Industrial Systems Engineering and Management developed a roadmap in which a series of indices can be tracked over time to quantify resilience and to monitor if it has improved over time. In particular, they conducted an empirical evaluation on the resilience of the U.S. power grid based on the database of the Electric Disturbance Events. Their work is currently published in the research journal, Reliability Engineering and System Safety.
In brief, the researchers commenced their empirical work by obtaining the Electric Disturbance Events (OE-417) power grid data. Next, to assess the trends in the systems resilience, they looked into three key components associated with each black-out and recovery of power systems. The three included: the time between disruptions, the performance loss of each disruption and the time needed for recovery. They then proposed a composite index and developed a modified Lewis–Robinson test to detect possible trends. Eventually, they performed trend tests on both the composite resilience index and in its three components in order to detect trends in the resilience of each electric region.
The authors were able to identify a period over which there was no trend in the occurrences of disruptions. The modified Lewis-Robinson test applied to the combined resilience measure suggested that the resilience in the Northeast Power Coordinating Council (NPCC) improves over the considered period. In addition, the test results showed a decreasing trend in the performance loss for the NPCC and the Reliability First (RF) regions, and decreasing trends in the normalized recovery time for the NPCC and the Midwest Reliability Organization (MRO) regions. Where results for the performance loss and the normalized recovery time were seen to tally with those for the resilience, a trend in resilience was detected. Overall, the resilience in the NPCC region was seen to have improved.
In summary, the study presented a comprehensive empirical evaluation on the resilience of the power grids of the North American Electric Reliability Corporation (NERC) regions based on the OE-417 data from Jan 2002 to Aug 2016. In general, they applied relevant trend tests in the context of resilience analysis. Furthermore, empirical evidence from the government financial support was used to substantiate their statistical findings. Altogether, their study presented a methodological approach that can be used in future to determine the reliability and resilience of power systems or other networking systems.
Lijuan Shen, Beatrice Cassottana, Loon Ching Tang. Statistical trend tests for resilience of power systems. Reliability Engineering and System Safety, volume 177 (2018) page 138–147.Go To Reliability Engineering and System Safety