Nowadays, power systems’ Protection, Automation, and Control (PAC) functionalities are often deployed in different constrained devices (Intelligent Electronic Devices) following a coupled hardware/software design. However, with the increase in distributed energy resources, more customized controllers will be required. These devices have high operational and deployment costs with long development, testing, and complex upgrade cycles. Addressing these challenges requires that a ’revolution’ in power system PAC design takes place. Decoupling from hardware-dependent implementations by virtualizing the functionalities facilitates the transition from a traditional power grid into a software-defined smart grid. This article presents a survey of recent literature on software-defined PAC for power systems, covering the concepts, main academic works, industrial proof of concepts, and the latest standardization efforts in this rising area. Finally, we summarize the expected future technical, industrial, and standardization challenges and open research problems. It was observed that software-defined PAC systems have a promising potential that can be leveraged for future PAC and smart grid developments. Moreover, standardizations in virtual IED software development and deployments, configuration tools, performance benchmarking, and compliance testing using a dynamic, agile approach assuring interoperability are critical enablers.

This article implements and validates the performance of a virtual intelligent electronic device (vIED) framework for digital substations using a real-time (RT) simulation environment. The work looks toward the future design of protection, automation, and control systems, an evolution of the digital substation design based on IEC 61850 and virtualization technology. An RT simulation setup was developed to speed up and enhance the deployment and maintenance of vIEDs with a novel well-defined test methodology. Several scenarios were tested by varying the number of vIEDs and relevant (communication, scalability, and functionality) configuration criteria. In addition, we assessed the efficiency of a software-defined communication network for the vIED framework and its adaptability to dynamic scaling of the network under transient data traffic loads. The tests demonstrated the efficient performance of vIEDs across various system configurations, particularly for requirements on the response time and network transfer latency. The findings showcase the significance of proper design and testing methodology that can be benchmarked against other virtualization platforms for substation systems.

Virtualization has emerged as a promising technological solution for electrical substations, driven by the convergence of information and operational technologies and the new requirements of cost efficiency and adaptability. Virtualization employs software to emulate hardware-level functionalities by means of an equivalent computing system. However, despite the promising potential that virtualization offers in driving cost efficiencies and enabling operational flexibility, there is a scarcity of techno-economic studies assessing its financial and societal impact. This paper proposes a novel, open-source cost-benefit assessment model based on net present value analysis (NPV) and system dynamics tailored to the case of protection, automation and control systems (PACs) in electrical substations. The model explores the capital expenditures (CAPEX) and operational expenditures (OPEX) of two alternatives: (1) traditional substations based on intelligent electronic devices (IEDs) and (2) virtualized PAC functions based on hardware-independent implementations deployed in virtual machines or containers. The economic model is evaluated for a 30-year simulation through comprehensive, optimistic and pessimistic scenarios of virtualization adoption as well as sensitivity analyses of relevant functional cases covering reliability, redundancy, initial costs, and virtualization roll-outs. The findings indicate that virtualization presents a promising technology for substations, leading to reductions in both CAPEX and update costs of approximately 20% and 60%, respectively, compared to traditional substations. Similarly, from a societal perspective, virtualization can offer positive key performance indicators associated with system downtime reductions and savings in power consumption and, consequently, in carbon footprint. However, the technology maturity, roll-out strategy, software licensing, and maintenance schemes can hinder the overall cost savings and affect payback periods.

Next-generation offshore wind farms are increasingly adopting vendor-agnostic software-defined networking (SDN) to oversee their Industrial Internet of Things Edge (IIoT-Edge) networks. The SDN-enabled IIoT-Edge networks present a promising solution for high availability and consistent performance-demanding environments such as offshore wind farm critical infrastructure monitoring, operation, and maintenance. Inevitably, these networks encounter stochastic failures such as random component malfunctions, software malfunctions, CPU overconsumption, and memory leakages. These stochastic failures result in intermittent network service interruptions, disrupting the real-time exchange of critical, latency-sensitive data essential for offshore wind farm operations. Given the criticality of data transfer in offshore wind farms, this paper investigates the dependability of the SDN-enabled IIoT-Edge networks amid the highlighted stochastic failures using a two-pronged approach to: (i) observe the transient behavior using a proof-of-concept simulation testbed and (ii) quantitatively assess the steady-state behavior using a probabilistic Homogeneous Continuous Time Markov Model (HCTMM) under varying failure and repair conditions. The study finds that network throughput decreases during failures in the transient behavior analysis. After quantitatively analyzing 15 case scenarios with varying failure and repair combinations, steady-state availability ranged from 93% to 98%, nearing the industry-standard SLA of 99.999%, guaranteeing up to 3 years of uninterrupted network service.