CITADELS: A Decentralized Approach to Grid-edge Coordination

With rapid decarbonization goals and ambitious urban electrification targets, the electric power distribution systems are undergoing unprecedented changes. The proliferation of distributed energy resources (DERs) and flexible loads, such as microgrids, roof-top and utility-scale photovoltaics, battery energy storage systems, electric vehicles, and grid-interfacing efficient buildings, is pushing the control and operational requirements of the grid to the edge. These grid-edge resources hold the potential to (1) enhance the grid’s operational flexibility to support decarbonization and electrification goals and (2) improve grid resilience in the aftermath of extreme weather events. However, massive penetrations of grid-edge resources with heterogeneous technologies, diverse user interaction modalities, and multi-ownership environments are significantly increasing the scale and complexity of grid operations, particularly at the edge, rendering traditional centralized architectures for grid communication and control ineffective. Effective use of grid-edge resources towards providing grid services necessitates new grid management strategies. The prevalent mechanisms include decentralized and distributed architecture for grid-edge coordination that, while being scalable, are also robust to single-point failures and can better manage data privacy considerations.

In a recently completed project, CITADELS[i] – led by Pacific Northwest National Laboratory (PNNL), Washington State University (WSU) collaborated with PNNL researchers to investigate new distributed/decentralized grid management strategies to increase the operational flexibility of the grid. This Grid Modernization Laboratory Consortium (GMLC) project was partly supported by the U.S. Department of Energy (DOE) under contract DE-AC05-76RL01830. The project team successfully developed, prototyped, and field-demonstrated the distributed/decentralized architectures to coordinate a network of microgrids to support the bulk power system during abnormal events and energize end-use loads when the bulk power systems fail. The successful completion of the CITADELS project serves as a proof of concept that the microgrid controllers can be coordinated using consensus algorithms on a secure pub/sub network (using the Open Field Message Bus approach), and a layered control architecture can be used to increase the grid’s operational flexibility by facilitating the control at the system “edge.” Note that at the edge, the team developed consensus algorithms to allow groups of microgrid controllers to communicate, exchange information, determine operational goals, and execute operational actions to achieve global objectives towards providing voltage and frequency support to the bulk grid during contingencies. To support the project goals, PNNL engaged numerous partners, which include the Electric Power Board of Chattanooga, Lawrence Livermore National Laboratory (LLNL), Oak Ridge National Laboratory (ORNL), Sandia National Laboratories (SNL), Commonwealth Edison (ComEd), Open Energy Solutions (OES), and WSU.

WSU researchers contributed to the CITADELS project by developing novel algorithms for the distributed coordination of microgrids to extract bulk-grid services from a large number of DERs. The proposed approach entailed a layered architecture for scalable DER coordination, including grid-forming and grid-following DERs for different grid services. The WSU team also developed a power-control-communication co-simulation platform using the Hierarchical Engine for Large-scale Infrastructure Co-Simulation (HELICS) to evaluate the performance of peer-to-peer and other distributed/decentralized grid-edge coordination algorithms. The proposed approaches were simulated and validated using the co-simulation platform on notional distribution systems based on IEEE distribution test feeders and utility distribution feeder models obtained from EPB. The team demonstrated the effectiveness of proposed distributed/decentralized coordination algorithms in providing voltage and frequency support. The co-simulation platform was also used to compare different distributed/decentralized coordination architectures for their convergence speed, accuracy in achieving the target objective, and robustness to communication non-idealities. This research helped evaluate the performance of advanced control architectures that heavily rely on communications and establish new value propositions for DERs/microgrids by enabling bulk-grid service provision.

The successful demonstration of distributed and decentralized control architectures in the CITADELS project opens new possibilities for grid-edge coordination. Although there are several advantages to distributed/decentralized coordination architectures, significant challenges remain to truly adopting these technologies. Currently, the deployment of distributed applications in an operational environment is in the early stages. The existing advanced distribution management systems are limited in their ability to host distributed applications. The limited understanding and non-standardized procedure for data transport among the distributed agents remains one of the crucial challenges in fully incorporating a distributed coordinated architecture in the utility operational paradigm. Moreover, several possibilities for communication and control hierarchies make it extremely challenging to develop a platform that can support different instantiations of distribution applications. Moving forward, further research is needed to explore the capabilities of layered control architecture along with the associated computing, and communication requirements for distribution systems. WSU researchers are currently collaborating with PNNL on a funded project, Grid Data Transport Analysis Framework (GDTAF), to answer some of these questions using the co-simulation platform. The goal is to quantitatively evaluate the value of data, computing, and communications and associated tradeoffs in achieving a desired level of distribution systems performance by coordinating/controlling grid-edge resources. Another initiative aims to develop and implement a layered architecture to coordinate diverse grid-edge resources effectively, enhancing the grid’s operational efficiency and resilience. The two primary projects where WSU researchers are collaborating closely with PNNL are: (1) the Spokane Connected Communities Project, funded by DOE, focused on developing a layered architecture for decision support from various grid-interactive buildings and distributed energy resources, and (2) RECUPERAT, funded by DOE’s Solar Energy Technology Office, aimed at establishing a layered architecture for automated restoration using PVs and storage after extreme weather events.

[i] CITADELS, Final Project Report, https://www.pnnl.gov/main/publications/external/technical_reports/PNNL-33889.pdf