SS10 What’s new in cascading failure analysis?

ABSTRACT This session will outline some of the latest advances in cascading failure analysis and resilience assessment of modern power systems. For many years, cascading failure analysis has largely been dominated by the use of static (power flow based) approaches as they provide huge computational benefits. However, these methods are increasingly coming under question as dynamic phenomena become more dominant in system failures. This session will cover a variety of techniques currently emerging at the forefront of cascading failure analysis and resilience assessment from dynamic modelling of the full system, the use of machine learning to predict cascading processes, and the consideration of probabilistic aspects to produce more accurate and meaningful statistics about the potential outcomes.

• Talk 1 (20 minutes, including direct Q&A): M. Panteli & S. Hashemi, University of Cyprus. Title: Integrated resilience and cascading modelling: quantification, mitigation and blackstart strategies This presentation will introduce novel cascading algorithms seamlessly integrated with spatial and temporal resilience analysis tools for quantifying the cascading impacts of large disturbances. It will then describe different mitigation strategies, such as preventive and corrective islanding while accounting for blackstart restoration, focusing on the critical decision-making on when and where to apply such strategies during the cascading propagation and restoration phase.
• Talk 2 (20 minutes, including direct Q&A): R. Preece, The University of Manchester Title: Benefits and challenges of dynamic modelling of cascading failures in power systems Time-based dynamic models of cascading failures have been recognized as one of the most comprehensive methods of representing detailed cascading information and are often used for benchmarking and validation. This talk will provide an overview of the progress in the field of dynamic analysis of cascading failures in power systems and outline the benefits and challenges of dynamic simulations in future grids. The benefits include the ability to capture temporal characteristics of system dynamics and provide timing information to facilitate control actions for blackout mitigation. The greatest barriers to dynamic modelling of cascading failures are the computational burden, and the extensive but often unavailable data requirements for dynamic representation of a power system.
• Talk 3 (20 minutes, including direct Q&A): G. jihad, Université Libre de Bruxelles. Title: Probabilistic dynamic methodologies for resilience assessment considering distributed energy resources. This talk will present probabilistic dynamic methodologies relevant to cascading failure analysis. Following this, their applicability will be demonstrated in the case of a resilience study by comparing the results obtained for a dynamic methodology to a static one.
• Talk 4 (20 minutes, including direct Q&A): P. Papadopoulos & T. Ahmad, University of Strathclyde. Title: Using machine learning to predict upcoming cascading events The evolution of cascading failures, especially at later stages before a system collapse are influenced by complex power system dynamics as well as the action of protection devices. This talk will discuss aspects related to modelling, sensitivity analysis, on the evolution of cascades as well as the use of machine learning (time-series based methods using LSTMs and Temporal Convolutional Networks) for predicting the onset and reason of upcoming cascading events and understanding the effect of topology (through Graph Convolutional Networks).
• General Q&A (10 minutes).

CHAIR

Dr Robin Preece, The University of Manchester

SPEAKERS

Robin Preece, The University of Manchester Mathaios Panteli, University of Cyprus Jihad Guenaou, Université Libre de Bruxelles Panagiotis Papadopoulos, University of Strathclyde Tabia Ahmad, University of Strathclyde

SHORT BIO

	Dr Robin Preece is a Reader in Future Power Systems in the Department of Electrical and Electronic Engineering at The University of Manchester, where he has been an academic since July 2014. Since then, he has helped to secure over £6 million in research funding for The University of Manchester. Dr Preece has published more than 90 international peer-reviewed papers in numerous different top-tier journals. His research is focussed on the dynamic stability of power systems with large quantities of power electronics and in quantifying the impacts of uncertainties and variability on network performance. He has presented his research at major international conferences hosted by the IET, IEEE, IFAC, and Cigré.
	Mathaios Panteli is currently an Assistant Professor with the Department of Electrical and Electronic Engineering, University of Cyprus. His main research interests include techno-economic reliability, resilience and flexibility assessment of future low-carbon energy systems, grid integration of renewable energy sources and integrated modelling and analysis of co-dependent critical infrastructures. Mathaios is an IEEE Senior Member, IET Chartered Engineer (CEng), the Chair of the CIGRE Working Group C4.47 “Power System Resilience” and CIGRE Cyprus National Committee and an active member of multiple IEEE working groups. He is also the recipient of the prestigious 2018 Newton Prize and was selected in the top 12 innovators for 2022 by the Innovation Radar Prize competition of the European Commission.
	Sina Hashemi currently works as a Post-Doctoral Research Fellow at the Department of Electrical and Computer Engineering, University of Cyprus. His current project is ‘Power System Black Start Restoration from Distributed Energy Resources (DERs)’. His research interest includes power system modelling, analysis, and control along with the application of machine learning and evolutionary optimization algorithms into electric power system issues.
	Jihad Guenaou is a teaching assistant and third year PhD student at Université Libre de Bruxelles in Brussels, Belgium. Her research explores the possibilities to use renewable energy resources to enhance power system resilience. This work combines transmission system analysis and the use of probabilistic dynamic methods to capture the impact and relevant dynamics of renewable energy resources.
	Panagiotis Papadopoulos is a Senior Lecturer and UKRI Future Leaders Fellow at the University of Strathclyde, working in the area of electric power systems. Panagiotis has received the Dipl. Eng. and Ph.D. degrees from Aristotle University of Thessaloniki, Greece, followed by a post-doctoral position at the University of Manchester. His research area is power system stability and dynamics and the application of data-driven methods and machine learning on power system online and offline dynamic security assessment. He has authored more than 60 papers in international journals and conferences and has been involved in 21 projects in collaboration with industrial partners.
	Tabia Ahmad is working as a Research Associate at the EEE department of University of Strathclyde, Glasgow (UK) working on the UKRI project, “Addressing the complexity of future power systems dynamic behaviour”. Her research interests include power system dynamics, WAMS based analytics, signal processing techniques in power systems and interpretable machine learning for power system applications. Prior to this she completed her Ph.D. thesis (with doctoral thesis distinction award) in electric power systems from the Indian Institute of Technology Delhi, India, and her BS and MS in Electrical Engineering, in 2014 and 2016, respectively, from India too