SEARCH · Search NASA
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
JADE provides a simple, flexible interface to submit jobs to an HPC. JADE is an HPC workflow management system that enables researchers to run simulations efficiently. It provides a simple, flexible interface to define jobs, as well as dependencies between jobs, and then dispatch them to an HPC with maximized parallelization. It also provides detailed error reporting and resource utilization statistics to aid in debugging problems.
A model structure to support representation of power system dispatching for optimal power flow that includes bid-in demands for load. Used to support the ARPA-E GO-3 competition.
Accurate energy forecasting is becoming critical due to many reasons: i ) optimal distributed energy resources operations and dispatch, ii) fault detection and diagnostics, and iii) meeting operational energy efficiency targets. Wattile uses deep learning (DL) for the building's short-term load forecasting application. Two specific types of neural networks called, Long Short Term Memory (LSTM) and Sequence-to-Sequence (S2S) models are used to make predictions. Forecasting models are trained using online historical weather and occupancy indicator data streams from the Intelligent Campus Program's data acquisition systems at the National Renewable Energy Laboratory (NREL) for main meters and sub-meters of multiple building types. These models use probabilistic methods to provide quantile-based forecasts in addition to nominal conditional median predictions of electricity consumption.
Development of Distributed Microgrid Controllers Using HELICS uses HELICS framework develop distributed microgrid controllers (DMCs) for coordinated grid service. Development of Distributed Microgrid Controllers Using HELICS includes thirteen agents, including OpenDSS agent, data manager agent, and eleven DMC agents. Development of Distributed Microgrid Controllers Using HELICS can be used as a standard framework to develop fully distributed control solutions that dispatches simulated DERs in simulation platform.
This repository contains code and data for performing a siting analysis of long-duration energy storage (LDES) in the 5-bus and RTS systems using the Sienna suite developed by the National Renewable Energy Laboratory for production cost modeling (PCM). This analysis involves moving the LDES component to different buses in the system, running a simulation, and considering the production cost of the simulation. Different system configurations are also analyzed with these scripts (such as moving load or renewable dispatch generators to different buses) to observe the impacts the system configuration has on optimal siting. This repository contains three sub directories discussed below. Both the 5-bus and RTS systems have two different initial configurations for the renewable energy components in them, one that is predominantly PV-driven and one that is predominantly wind-driven. Sienna suite can be found here: https://github.com/NREL-Sienna
This codebase contains a multi-agent, actor-critic reinforcement learning implementation for cyber-resilient grid voltage control. It uses a 123-bus OpenDSS system as the environment, with three-phase power flow translating nodal power injections into solved nodal voltages. The reward function penalizes deviations from nominal voltage as well as reactive power dispatch, while encouraging agents to take actions that result in fast convergence to nominal conditions. The codebase models false data injection attacks and includes functionality for training, testing, hyper-parameter tuning, and visualization.
Data and a supporting lightweight Python package that describes possible costs for enabling EV managed charging from 2025 to 2050 for three dispatch mechanisms: Time-of-Use (TOU), Real Time Pricing (RTP), and direct load control (DLC) and four flexibility scenarios (Flat and Low, Mid, and High Flex).
The new version of HydraGNN v4.0 provides additional core capabilities, such as: Inclusion of multi-body atomistic cluster expansion MACE, polarizable atom interaction neural network PAINN, and equivariant principal neighborhood aggregation (PNAEq) among the message passing layers supported -Inclusion of graph transformers to directly model long-range interactions between nodes that are distant in the graph topology Integration of graph transformers with message passing layers by combining the graph embedding generated by the two mechanisms, which allows for an improved expressivity of the HydraGNN architecture Improved re-implementation of multi-task learning (MTL) to allow its use for stabilized training across imbalanced, multi-source, multi-fidelity data Introduction of multi-task parallelism, a newly proposed type of model parallelism specifically for MTL architectures, which allows to dispatch different output decoding heads to different GPU devices Integration of multi-task parallelism with pre-existing distributed data parallelism to enable a 2D parallelization for distributed training Improved portability of the distributed training across Intel GPUs, which has been testes on ALCF exascale supercomputer Aurora Inclusion of 2-level fine-grained energy profilers portable across NVIDIA, AMD, and Intel GPUs to monitor the power and energy consumption associated with different functions executed by the HydraGNN code during data pre-load and training Restructuring of previous examples and inclusion of new sets of examples to illustrate the download, preprocess, and training of HydraGNN models on new large-scale open-source datasets for atomistic materials modeling (e.g., Alexandria, Transition1x, OMat24, OMol25)
SAND2025-11838O REDLY provides a framework for designing, training, and validating physics-informed neural networks for power system applications. It trains surrogates for Alternating Current Power Flow prediction and general economic dispatch problems, such as Alternating Current Optimal Power Flow and Direct Current Power Flow. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.
SAND2025-11463O The Real Vector Framework (RVF) is a modern and flexible C++ vector math library for developing scientific computing software that involves vector computations. RVF allows an opt-in approach to functionality that parallels the familiar base-class and override structures of object-oriented programming. Users can reuse and customize the code without inheritance entanglements and dynamic dispatch, while enabling seamless interoperability between diverse container types. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.
The Automated Control Testbed for Integration, Verification, and Emulation (ACTIVE) framework is a software platform designed to support the optimized operation and management of a wide range of building types. It enables the development, testing, and validation of diverse control strategies, including AI-based, rule-based, and model-based approaches. The platform facilitates a seamless transition from simulation-based evaluation of control strategies to real-world field validation and deployment. ACTIVE supports the full building management lifecycle, encompassing data acquisition and management, system monitoring, optimized control, adaptive learning services, device dispatch and coordination, as well as advanced analytics and visualization. Together, these capabilities provide an integrated environment for improving building performance, operational efficiency, reducing energy cost, and reliability.
matsim-agents is a multi-agent AI framework for atomistic materials simulation and discovery. It orchestrates large language models (LLMs), machine-learned interatomic potentials (MLIPs), and DFT codes into a single agentic loop running on laptops and DOE leadership-class supercomputers. MULTI-AGENT ORCHESTRATION A LangGraph state machine with three nodes: a Planner that converts a natural-language research objective into structured tasks; an Executor that dispatches atomistic tools and loops until the queue is empty; and an Analyst that summarizes results into a human-readable report. State is checkpointed after every step and human-in-the-loop gates can be inserted at any edge. HYPOTHESIS-DRIVEN DISCOVERY CHAT An interactive REPL (matsim-agents chat) that couples LLM dialogue with atomistic simulation. Chemical formulas are automatically detected in conversation turns and trigger a full crystal-phase exploration: structure generation → relaxation → stability scoring → result injection back into the conversation, creating a closed hypothesis-refinement loop. CRYSTAL PHASE ENUMERATION Given a composition, the phase explorer enumerates prototypes by stoichiometry: elemental (fcc/bcc/hcp/sc/diamond), binary 1:1 (rocksalt/CsCl/zincblende/ wurtzite/fluorite/rutile), ternary 1:1:3 (cubic perovskite), ternary 1:2:4 (perovskite + spinel), quaternary 1:1:2:6 (Fm-3m double perovskite). 2-D prototypes (graphene, h-BN, MoS2 2H/1T) and multilayer stacking are also supported via --include-2d and --num-layers. SUPERCELL GENERATION AND SITE DECORATION Auto-tiling to a minimum atom count (--min-atoms), explicit NxNxN tiling (--supercell), symmetry-distinct site decorations (--n-orderings), and isotropic lattice-scale sweeps (--lattice-scales) for volume bracketing. MLFF RELAXATION AND STABILITY SCORING HydraGNN (multi-headed GNN) drives structure relaxation via ASE with FIRE, BFGS, or BFGSLineSearch. Stability output: delta-E/atom ranking across phases and a max-residual-force dynamical-stability proxy. Other MLIPs (MACE, NequIP, Orb) can be plugged in through the same interface. DFT BACKENDS Quantum ESPRESSO pw.x and VASP 6.6 are first-class labellers. Both have validated GPU builds and SLURM/PBS launchers for three DOE platforms: Frontier (AMD MI250X, ROCm), Aurora (Intel PVC, oneAPI), Perlmutter (NVIDIA A100, CUDA). QE produces ~100 binaries (pw.x, ph.x, epw.x, ...). VASP supports scf, relax, vc-relax, and vc-relax-shape run types. ACTIVE-LEARNING LOOP matsim-agents al run CONFIG.yaml drives an iterative HydraGNN-DFT loop: MD generates candidates → ensemble/MC-dropout uncertainty selects the most informative → DFT labels them in parallel inside one allocation → dataset grows → HydraGNN retrains → repeat. DFT backend is a single YAML toggle (dft.backend: vasp | qe). LLM-generated seed structures are supported (no curated POSCAR library needed). Config uses ${VAR}, ${VAR:-default}, ${VAR:?msg} shell-style substitution for cross-user/cross-site portability. LLM BACKENDS Ollama (local, default), vLLM (HPC multi-GPU serving), OpenAI, Anthropic, HuggingFace Transformers+Accelerate. Selected at runtime via flag or env var with no code changes. HPC PORTABILITY Same Python entry points run on Frontier (ROCm 7.2), Aurora (oneAPI), and Perlmutter (CUDA 12). DFT and ML stacks are never co-loaded in the same shell; they couple through the scheduler and filesystem. Advanced multi-node launchers (serve, discovery-chat, single-relaxation, active-learning, QE warm-start) are provided for all three platforms. CODABENCH COMPETITION BUNDLE A self-contained benchmark: 159 atomistic test structures across 11 material classes, 5 tasks (formation energy, forces, ML relaxation, AI-DFT relaxation, phase stability ranking), public/private leaderboard split (30/70), and four ready-to-run baselines: MACE-MP-0, HydraGNN, UMA, AllScAIP.
AQDrop is a job management system designed to streamline access to the Advanced Quantum Testbed (AQT) at NERSC (National Energy Research Scientific Computing Center). It serves as a centralized middleware layer between researchers and quantum processing hardware. Key Features: AQDrop provides a FastAPI-based server backed by PostgreSQL for job submission, queue management, and role-based access control (members, operators, and administrators). Users submit Qiskit circuits via JSON payloads, which are queued, dispatched to the QPU through the Qubic API, and returned as measurement counts. A Python client library and web dashboard round out the interface options. Primary Use: Researchers submit quantum circuit jobs from a laptop or login node; an operator client executes those jobs on the AQT's physical QPU and returns results — all coordinated through the central API. Advantages: Compared to ad-hoc or direct hardware access, AQDrop adds structured queue management, auditable job-status tracking and OAuth2 authentication — reducing scheduling conflicts and unauthorized access. Its containerized deployment also improves reproducibility and scalability. Overall, AQDrop functions as a purpose-built quantum job broker tailored to NERSC's specific hardware and institutional access requirements.
Grid Analysis Toolkit (GAT) is a unified Python API and plotting for power system PCM and CEM results (Sienna, PLEXOS, ReEDS™). It's a toolkit for wrangling data for Bulk Grid Dispatch and Transmission Analysis. GAT aims to provide simplified access to PCM and CEM results in a standard format while also allowing raw data access to underlying datasets specific to the model. This software can also be found on PyPI at For plotting, GAT defaults to standard National Lab of the Rockies (NLR) color schemes and standard styles while allowing customization.
Southern Company is an electric utility that is embracing residential microgrids as a platform for distributed energy. The first residential microgrid in the southeastern United States was constructed in Hoover, Alabama, as a testbed for development. To operate the microgrid, we developed an optimization model for unit commitment and economic dispatch. We report the optimization model was implemented in a microgrid central controller and tested under various operational scenarios. Challenges also include developing a pricing algorithm to deliver price forecasts to the residential homes in the microgrid and open-source software restrictions.
This dataset comprises temporal dynamic graph sequences generated from power grid simulations focused on grid reconfiguration to enhance resilience. The simulations model failure propagation under varying conditions, with nodes assigned distinct failure probabilities. For each time step, the dataset captures the evolution of node states (functional or failed) and features critical to grid operations, such as pv_output, load_profile, load_dispatch, dg_output, loss, and voltage. Node types include sources, normal loads, and nodes with specific equipment like PVs, micro turbines, or shunt capacitors. The dataset is structured to support the training of dynamic graph neural networks, facilitating research on node feature prediction and edge dynamics under failure scenarios. Three distinct configurations are included, providing a robust foundation for modeling power grid resilience.
Concentrating solar power (CSP) coupled with thermal energy storage (TES) is being considered as an appealing solution to deliver stable, dispatchable, and inexpensive electricity generation from renewable solar energy. The heat from the HTF is stored in a TES system that is significantly less expensive and more scalable than electrochemical storage, enabling electricity generation when the sunlight is unavailable. The general concentrating solar thermal technologies are being pushed to higher temperature for higher efficiency and lower cost in a diverse range application in electricity generation (in CSP), industrial heating, and thermochemical fuel productions. Inert and abundant solid particles are becoming increasingly more important for these emerging applications. Heat transfer in both stationary and moving particle beds have been studied for many decades, revealing a great deal of insights on the mechanisms. Yet, there are still several unresolved problems from both fundamental understanding and technological applications points of view. In this Chapter, we will attempt to provide an up-to-date picture of thermal transport in stationary and moving particle beds. In both cases, we will review historical and contemporary developments in the theoretical, computational, and experimental aspects of heat transfer and related behaviors in particle beds. Here, we will discuss both the advancements and limitations of the existing studies and point to future developments towards deeper fundamental understandings and better engineering applications.
Thermal energy storage (TES) deployed with concentrating solar power (CSP) has shown value and capabilities in generating dispatchable power. It has attracted significant interest for potential roles in energy storage for various capacities and applications. Energy storage plays an important role in expanding renewable generation and is becoming indispensable in future carbon-free electric power. TES development and deployment provide various possibilities for electric storage and have shown advantages in capacity and flexibility to support renewable integration. In addition, renewable energy integrated with a low-cost TES system can potentially provide a reliable thermal energy supply to displace fossil fuels for broad decarbonization. Varieties of TES technologies and developments breed new areas of growth in TES materials, designs, components, and system configurations, which provide opportunities for substantial cost reduction and performance improvements to meet the technoeconomic challenges in practical applications. This chapter will describe recent progress in thermal energy storage research and development and primarily focus on sensible heat storage and its applications in supplying reliable electricity and industry process heat by integrating with variable renewable energy.