Critical infrastructure sectors are becoming increasingly complex and interdependent and so, too, are the physical and socioeconomic landscapes in which they operate. This in turn has increased the possibility of disruptions and subsequent compounding disruptions. Climate change adds an additional layer of uncertainty and disruption to these evolving interconnected networks. And, for the power grid, this shifting threat landscape is exacerbated by increased demand from electrification, decarbonization, population growth, and an overall increase in reliance on electricity. Climate change is also affecting energy usage patterns. For example, Texas has seen spikes in energy demand and intense strain on the power grid from both unexpected cold snaps and intense heat waves.
The increase in electric demand is not occurring in isolation. The grid edge is also changing. These changes are in turn stressing the grid from two angles: capacity and control. The first, capacity, falls under the purview of NERC and is regularly studied and planned for through Reliability Assessments.
However, the second, the ability to control and make real-time decisions in a grid that is increasingly physically decentralizing and relying on weather-dependent distributed energy resources (DERs), does not have an established guiding principle. Leveraging resilience-based principles for multi-agent control mechanisms can improve the power grid’s ability to meet society’s needs despite these added stressors.
Multi-agent decision-making and control is a planning and operational approach that acknowledges the autonomous and bi-directional nature of the additional energy resources needed to meet our capacity demands in an environmentally friendly and equitable manner. Because the power grid is decentralizing, a bottom-up tactic for maintaining a reliable and resilient network ensures that the increased DERs, intelligent energy devices, and advanced controls necessary to meet increased demand is computationally possible.[1] Of course, this requires some software upgrades to our existing power grid.
Digitalization and machine learning (ML) provide the backbone to these upgrades. Digitalization has three dimensions or pillars: Converting the physical to the digital, implementing digital tools to analyze the data, and transforming the data into decision making. ML is the use of programs or models that leverage computer algorithms and statistics to find patterns in data, make predictions, and advance the decision-making protocols. In other words, digitalization is comprised of data, analytics and digital transformation, of which ML is one tool for analyzing the data and transforming it into business or operational decisions.
So far, digitalization and ML have enabled utilities and regulatory bodies, such as independent system operators (ISOs), to increase real-time monitoring and decision-making for improved operations and planning. This has been spurred by the rapid adoption of sensors, and data storing and transmitting technologies, that have been placed throughout the power grid network due to their affordability and information sharing capabilities. This gives us a better understanding of our equipment and the surrounding conditions, in addition to facilitating demand-response programs that integrate end-users into the network for improved system performance and reduced emissions.
Also, with digitalization comes digital twins, or computerized versions of the system, further enhancing our ability to make informed system-wide decisions. Loads can be better understood, and predictions on how the system will function throughout the disconnection process of a power line or substation can be made, giving us better predictions on how the network will function during critical moments. These simulations provide powerful tools for forecasting energy production, asset management and system-wide planning.
The power grid is a critical system that is based on legacy infrastructure but undergoing a rapid digital transformation. Compounded with the interconnected systems and infrastructure relying on resilient power, electrification and decarbonization driving power grid expansion, the changing grid edge (e.g., residential solar, EVs, batteries, etc.), and climate change, the power grid is at a crossroads.
In addition to investments in expanding our bulk power transfer, the future grid should facilitate and encourage local energy sourcing to optimize the utilization of DER and provide decentralized resilience. It should also incorporate advanced control software (i.e., ML-based solution) to equip system operators with tools for faster decision-making.
[1]https://www.nrel.gov/news/features/2019/from-the-bottom-up-designing-a-decentralized-power-system.html
Maureen S. Golan is a Ph.D. candidate in the Department of Civil, Architectural, and Environmental Engineering at the University of Texas at Austin, researching power grid resilience and effective quantification methods. She has a Master of Science in Civil and Environmental Engineering from Carnegie Mellon University and is a returned Peace Corps volunteer (Vanuatu). She can be reached at mgolan@utexas.edu.
- Maureen S. Golanhttps://energiesmedia.com/author/maureen-s-golan/
Oil and gas operations are commonly found in remote locations far from company headquarters. Now, it's possible to monitor pump operations, collate and analyze seismic data, and track employees around the world from almost anywhere. Whether employees are in the office or in the field, the internet and related applications enable a greater multidirectional flow of information – and control – than ever before.