David Chassin of Eudoxys Sciences’ presentation at LF Energy Summit 2024 provided key insights into the development, challenges, and future of Arras Energy, an open source tool designed to support research and simulation for the evolving energy grid. A summary of the session follows, and the full video can be found at the end of this post.
History and Development
Arras Energy, the Linux Foundation’s commercialized version of a U.S. Department of Energy (DOE) tool, has a 25-year history rooted in collaboration with various research institutions, including Stanford University, Pacific Northwest National Laboratory (PNNL), and others. Initially developed for research purposes under the DOE’s GridLab-D, it was designed to solve complex grid-related problems and help model future scenarios for the grid, such as extreme weather events, deep electrification, and tariff studies.
Although originally designed for research rather than high-performance utility use, the tool has evolved to address larger-scale utility needs. With support from organizations like the California Energy Commission and collaboration with major partners such as Amazon Web Services (AWS), the tool has undergone significant testing and fielding at various utilities, including National Grid.
Key Capabilities and Use Cases
Arras Energy is a powerful simulation engine, providing solutions to grid planning and operation problems that traditional tools struggle with. Its high-performance capabilities enable utilities to model scenarios 20 years into the future, a critical function for modernizing the energy grid. Some key use cases include:
- Hosting Capacity Analysis: Modeling the integration of distributed energy resources (DERs) and assessing grid hosting capacity for these assets.
- Extreme Event Simulations: Understanding how the grid can withstand and recover from extreme weather events or other system shocks.
- Deep Electrification: Exploring the impact of significant electrification, such as electric vehicles and distributed storage, on grid stability and performance.
- Alternative Tariffs: Analyzing the impact of various pricing models on grid operations and customer behavior.
Chassin emphasized that the tool allows flexibility in the methods used, integrating Python for event handling and machine learning for solving power flow problems.
Challenges in Open Source
Despite its origins in open source, transitioning the tool from a DOE project to a commercialized open source platform under LF Energy has not been without hurdles. The following challenges stand out:
- Open Source Complexities: Open source often presents unexpected complexities, from maintaining dependencies to ensuring compatibility across updates. Arras Energy’s reliance on a wide variety of modules and contributors means that updates can break the system, making maintenance labor-intensive.
- Licensing and Data Providence: The open source nature of the tool brings licensing challenges, especially around code and data provenance. Managing data licenses is critical, and many organizations still misunderstand open source, equating it with “free,” when in reality it requires significant expertise and resources to deploy and maintain.
- User Interfaces and Accessibility: Several user interfaces have been developed for Arras Energy, including Hitachi’s Glow UI, to improve accessibility and usability. However, Chassin highlighted the need for better documentation and community-building efforts to make the tool more approachable for users, particularly in large-scale utility settings.
The Future of Arras Energy
Looking ahead, Chassin believes that the future of Arras Energy lies in solving the key problem of balancing flexibility and stability within open source environments. As the energy landscape continues to evolve, the tool must adapt to manage the increasing complexities of grid modernization, all while maintaining a scalable, open source foundation. He also pointed to the potential of machine learning to enhance the tool’s performance, making it an invaluable resource for grid operators and researchers alike.
