Search NASA⌕ Search

SEARCH · Search NASA

Results for “explainable AI”

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.

At least 181 records · Page 10
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Surfactant-Specific AI-Driven Molecular Design: Integrating Generative Models, Predictive Modeling, and Reinforcement Learning for Tailored Surfactant Synthesis

Molecular design is a critical aspect of various scientific and industrial fields, where the properties of molecules hold significant importance. In this study, a 3-fold methodology design is presented that leverages the power of generative artificial intelligence (AI), predictive modeling, and reinforcement learning to create tailored molecules with desired properties. This model synergistically combines deep learning techniques with Self-Referencing Embedded Strings (SELFIES) molecular representation to build a generative model that generates valid molecules and a graphical neural network model that accurately forecasts molecular properties. The Variational Autoencoder (VAE) coupled with reinforcement learning helps refine molecule generation based on targeted attributes. Data from an experimental study involving surfactants were used to test the framework. A validation of the structural integrity of the molecules generated was conducted, and Tanimoto similarities were used to quantify the similarity and diversity between the original and generated molecular structures. Also, saliency maps for the generated surfactants were produced to identify the features explaining the property values. Lastly, molecular dynamics simulations were used to validate the stability of the generated molecules. The results showed that the proposed framework can effectively produce valid molecules within the set property threshold value.

36 MATERIALS SCIENCE↗
DOE OSTI2016

The universal “heartbeat” oscillations in black hole systems across the mass-scale

The hyperluminous X-ray source (HLX-1, peak X-ray luminosity ∼10{sup 42} erg s{sup −1}) near the spiral galaxy ESO 243-49, which underwent recurrent outbursts within a period of ∼400 days, is possibly the best candidate for an intermediate mass black hole (IMBH). The physical reason for this quasiperiodic variability is still unclear. We explore the possibility of radiation-pressure instability in the accretion disk by modeling the light curve of HLX-1, and find that it can roughly reproduce the duration, period, and amplitude of the recurrent outbursts in HLX-1 with an IMBH of ∼10{sup 5} M{sub ⊙}. Our result provides a possible mechanism to explain the recurrent outbursts in HLX-1. We further find a universal correlation between the outburst duration and the bolometric luminosity for the black hole (BH) sources with a very broad mass range (e.g., X-ray binaries, HLX-1, and active galactic nuclei), which is roughly consistent with the prediction of radiation-pressure instability of the accretion disk. These results imply that “heartbeat” oscillations triggered by radiation-pressure instability may appear in different-scale BH systems.

79 ASTRONOMY AND ASTROPHYSICS↗
NASA NTRS2011

Signatures of Chemical Evolution in Protostellar Nebulae

A decade ago observers began to take serious notice of the presence of crystalline silicate grains in the dust flowing away from some comets. While crystallinity had been seen in such objects previously, starting with the recognitions by Campins and Ryan (1990) that the 10 micron feature of Comet Halley resembled that of the mineral forsterite, most such observations were either ignored or dismissed as no path to explain such crystalline grains was available in the literature. When it was first suggested that an outward flow must be present to carry annealed silicate grains from the innermost regions of the Solar Nebula out to the regions where comets could form (Nuth, 1999; 2001) this suggestion was also dismissed because no such transport mechanism was known at the time. Since then not only have new models of nebular dynamics demonstrated the reality of long distance outward transport (Ciesla, 2007; 2008; 2009) but examination of older models (Boss, 2004) showed that such transport had been present but had gone unrecognized for many years. The most unassailable evidence for outward nebular transport came with the return of the Stardust samples from Comet Wild2, a Kuiper-belt comet that contained micron-scale grains of high temperature minerals resembling the Calcium-Aluminum Inclusions found in primitive meteorites (Zolensky et aI., 2006) that formed at T > 1400K. Now that outward transport in protostellar nebulae has been firmly established, a re-examination of its consequences for nebular gas is in order that takes into account both the factors that regulate both the outward flow as well as those that likely control the chemical composition of the gas. Laboratory studies of surface catalyzed reactions suggest that a trend toward more highly reduced carbon and nitrogen compounds in the gas phase should be correlated with a general increase in the crystallinity of the dust (Nuth et aI., 2000), but is such a trend actually observable? Unlike the Fischer-Tropsch or the Haber-Bosch reactions used in industry, the surface catalyzed reactions seen in our laboratory do not produce a simple product stream of methane or ammonia, respectively. Instead, such reactions produce a wide range of both aliphatic and aromatic hydrocarbons, as well as reduced nitrogen compounds such as ammonia, amines, amides and imides, as gas phase products together with a heavy, macromolecular, kerogen-like surface coating that remains on the grains. While CO and N2 will certainly be depleted by conversion into more complex and less volatile species via reaction on grain surfaces, it may be very difficult to monitor such changes from outside the system.

Nuth, Joseph A., III↗
DOE OSTI2018

Transient Spectroscopy of Glass-Embedded Perovskite Quantum Dots: Novel Structures in an Old Wrapping

Semiconductor doped glasses had been used by the research and engineering communities as color filters or saturable absorbers well before it was realized that their optical properties were defined by tiny specs of semiconductor matter known presently as quantum dots (QDs). Nowadays, the preferred type of QD samples are colloidal particles typically fabricated via organometallic chemical routines that allow for exquisite control of QD morphology, composition and surface properties. However, there is still a number of applications that would benefit from the availability of high-quality glass-based QD samples. These prospective applications include fiber optics, optically pumped lasers and amplifiers and luminescent solar concentrators (LSCs). In addition to being perfect optical materials, glass matrices could help enhance stability of QDs by isolating them from the environment and improving heat exchange with the outside medium. We conduct optical studies of a new type of all-inorganic CsPbBr 3 perovskite QDs fabricated directly in glasses by high-temperature precipitation. These samples are virtually scattering free and exhibit excellent waveguiding properties which makes them well suited for applications in, for example, fiber optics and LSCs. However, the presently existing problem is their fairly low room-temperature emission quantum yields of only ca. 1%–2%. Here we investigate the reasons underlying the limited emissivity of these samples by conducting transient photoluminescence (PL) and absorption measurements across a range of temperatures from 20 to 300K. We observe that the low-temperature PL quantum yield of these samples can be as high as ~25%. However, it quickly drops (in a nearly linear fashion) with increasing temperature. Interestingly, contrary to traditional thermal quenching models, experimental observations cannot be explained in terms of a thermally activated nonradiative rate but rather suggest the existence of two distinct QD sub-ensembles of “emissive” and completely “nonemissive” particles. The temperature-induced variation in the PL efficiency is likely due to a structural transformation of the QD surfaces or interior leading to formation of extremely fast trapping sites or nonemissive phases resulting in conversion of emissive QDs into nonemissive. Thus, future efforts on improving emissivity of glass-based perovskite QD samples might focus on approaches for extending the range of stability of the low-temperature highly emissive structure/phase of the QDs up to room temperature.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRS1988

Expert system validation in prolog

An overview of the Expert System Validation Assistant (EVA) is being implemented in Prolog at the Lockheed AI Center. Prolog was chosen to facilitate rapid prototyping of the structure and logic checkers and since February 1987, we have implemented code to check for irrelevance, subsumption, duplication, deadends, unreachability, and cycles. The architecture chosen is extremely flexible and expansible, yet concise and complementary with the normal interactive style of Prolog. The foundation of the system is in the connection graph representation. Rules and facts are modeled as nodes in the graph and arcs indicate common patterns between rules. The basic activity of the validation system is then a traversal of the connection graph, searching for various patterns the system recognizes as erroneous. To aid in specifying these patterns, a metalanguage is developed, providing the user with the basic facilities required to reason about the expert system. Using the metalanguage, the user can, for example, give the Prolog inference engine the goal of finding inconsistent conclusions among the rules, and Prolog will search the graph intantiations which can match the definition of inconsistency. Examples of code for some of the checkers are provided and the algorithms explained. Technical highlights include automatic construction of a connection graph, demonstration of the use of metalanguage, the A* algorithm modified to detect all unique cycles, general-purpose stacks in Prolog, and a general-purpose database browser with pattern completion.

Stock, Todd↗
DOE OSTI2025

LC-Opt: Benchmarking Reinforcement Learning and Agentic AI for End-to-End Liquid Cooling Optimization in Data Centers

Liquid cooling is critical for thermal management in high-density data centers with the rising AI workloads. However, machine learning-based controllers are essential to unlock greater energy efficiency and reliability, promoting sustainability. We present LC-Opt, a Sustainable Liquid Cooling (LC) benchmark environment, for reinforcement learning (RL) control strategies in energy-efficient liquid cooling of high-performance computing (HPC) systems. Built on the baseline of a high-fidelity digital twin of Oak Ridge National Lab's Frontier Supercomputer cooling system, LC-Opt provides detailed Modelica-based end-to-end models spanning site-level cooling towers to data center cabinets and server blade groups. RL agents optimize critical thermal controls like liquid supply temperature, flow rate, and granular valve actuation at the IT cabinet level, as well as cooling tower (CT) setpoints through a Gymnasium interface, with dynamic changes in workloads. This environment creates a multi-objective real-time optimization challenge balancing local thermal regulation and global energy efficiency, and also supports additional components like a heat recovery unit (HRU). We benchmark centralized and decentralized multi-agent RL approaches, demonstrate policy distillation into decision and regression trees for interpretable control, and explore LLM-based methods that explain control actions in natural language through an agentic mesh architecture designed to foster user trust and simplify system management. LC-Opt democratizes access to detailed, customizable liquid cooling models, enabling the ML community, operators, and vendors to develop sustainable data center liquid cooling control solutions.

Naug, Avisek [Hewlett Packard Enterprise]↗
DOE OSTI2023

Progress on the experimental search for the chiral magnetic effect, the chiral vortical effect, and the chiral magnetic wave

In quantum chromodynamics, the interactions of quarks with the topological gluon field can lead to nonconservation of local parity (P) and conjugated parity (CP) , which provides a solution to the strong CP problem and a possibility to explain the asymmetry of matter-antimatter in the current universe. Under the action of a strong magnetic field, the nonconservation of P and CP can lead to the separation of particles according to their electric charges, which is called the chiral magnetic effect (CME). An observation of the CME-induced charge separation will confirm several fundamental properties of quantum chromodynamics (QCD), namely, approximate chiral symmetry restoration, topological charge fluctuation, and local parity violation. In relativistic heavy-ion collisions, there are other chiral anomalous effects similar to the CME, such as the chiral vortical effect (CVE) and the chiral magnetic wave (CMW). This review briefly summarizes the current progress of experimental research on the CME, CVE, and CMW in relativistic heavy-ion collisions.

Physics↗
DOE OSTI2023

Performance Comparison of Object Detection Networks for Shrapnel Identification in Ultrasound Images

Ultrasound imaging is a critical tool for triaging and diagnosing subjects but only if images can be properly interpreted. Unfortunately, in remote or military medicine situations, the expertise to interpret images can be lacking. Machine-learning image interpretation models that are explainable to the end user and deployable in real time with ultrasound equipment have the potential to solve this problem. We have previously shown how a YOLOv3 (You Only Look Once) object detection algorithm can be used for tracking shrapnel, artery, vein, and nerve fiber bundle features in a tissue phantom. However, real-time implementation of an object detection model requires optimizing model inference time. Here, we compare the performance of five different object detection deep-learning models with varying architectures and trainable parameters to determine which model is most suitable for this shrapnel-tracking ultrasound image application. We used a dataset of more than 16,000 ultrasound images from gelatin tissue phantoms containing artery, vein, nerve fiber, and shrapnel features for training and evaluating each model. Every object detection model surpassed 0.85 mean average precision except for the detection transformer model. Overall, the YOLOv7tiny model had the higher mean average precision and quickest inference time, making it the obvious model choice for this ultrasound imaging application. Other object detection models were overfitting the data as was determined by lower testing performance compared with higher training performance. In summary, the YOLOv7tiny object detection model had the best mean average precision and inference time and was selected as optimal for this application. Next steps will implement this object detection algorithm for real-time applications, an important next step in translating AI models for emergency and military medicine.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2024

Learning dynamical systems from data: An introduction to physics-guided deep learning

Modeling complex physical dynamics is a fundamental task in science and engineering. Traditional physics-based models are first-principled, explainable, and sample-efficient. However, they often rely on strong modeling assumptions and expensive numerical integration, requiring significant computational resources and domain expertise. While deep learning (DL) provides efficient alternatives for modeling complex dynamics, they require a large amount of labeled training data. Furthermore, its predictions may disobey the governing physical laws and are difficult to interpret. Physics-guided DL aims to integrate first-principled physical knowledge into data-driven methods. It has the best of both worlds and is well equipped to better solve scientific problems. Recently, this field has gained great progress and has drawn considerable interest across discipline Here, we introduce the framework of physics-guided DL with a special emphasis on learning dynamical systems. We describe the learning pipeline and categorize state-of-the-art methods under this framework. We also offer our perspectives on the open challenges and emerging opportunities.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Hierarchical Multi-agent Large Language Model Reasoning for Autonomous Heterogeneous Catalyst Discovery

Artificial intelligence is reshaping scientific exploration, but most methods automate procedural tasks without engaging in scientific reasoning, limiting autonomy in discovery. We demonstrate that hierarchical agentic large language model reasoning can efficiently drive simulation and scientific exploration. Across two chemical applications, CO adsorption on Cu surface transition metal adatoms and on M–N–C catalysts, reasoning-guided exploration reduces required atomistic simulations by up to 90% relative to heuristic or random selection. Comparisons across single-agent, multi-agent, and stochastic baselines show that hierarchical strategies yield more coherent and information-efficient search trajectories. Reasoning traces reveal chemically grounded decisions that cannot be explained by semantic bias or stochastic sampling. We realize these agentic reasoning strategies in Materials Agents for Simulation and Theory in Electronic-structure Reasoning (MASTER), a multimodal system that translates natural language into density functional theory workflows. Altogether, multi-agent collaboration accelerates heterogeneous catalyst discovery and marks a step toward more autonomous, reasoning-guided scientific exploration.

30 DIRECT ENERGY CONVERSION↗
DOE OSTI2020

Methods for Explainable Artificial Intelligence

We explored ways of quantifying information in a neural network. This can be used to determine the right size of a network or to infer the way in which a network is processing information. The first year and a half was somewhat exploratory while the last half of the project focused on approaches that seemed to show the most promise. The introduce a new way of computing explainable artificial intelligence (XAI) saliency maps that is several orders of magnitude faster than methods with similar fidelity We call it FastCAM. The method works be combining a Class Activation Map (CAM) method such as GradCAM with a forward activation map computed with a statistic we call SMOE Scale. The addition of the forward activation maps to CAM methods seems to always improve their fidelity. At the same time, computational overhead is not increased by very much. While Gradients with SmoothGrad scores better on some fidelity measures, it is overall not as good and requires more than 1500 times to compute. We demonstrate two completed applications of FastCAM on tasks outside of the LDRD at LLNL. The source code for FastCAM is currently being implanted into Captum, the official XAI toolkit for the popular deep learning toolkit PyTorch. The LDRD currently has 13 publications released to the public. 10 of them are journal length.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Explaining Nonmerger Gamma-Ray Bursts and Broad-lined Supernovae with Close Binary Progenitors with Black Hole Central Engines

For over 25 yr, the origin of long-duration gamma-ray bursts (lGRBs) has been linked to the collapse of rotating massive stars. However, we have yet to pinpoint the stellar progenitor powering these transients. Moreover, the dominant engine powering the explosions remains open to debate. Observations of both lGRBs, supernovae associated with these GRBs, such as broad-line (BL) stripped-envelope (type Ic) supernovae (hereafter, Ic-BL), supernovae (SNe), and perhaps superluminous SNe, fast blue optical transients, and fast x-ray transients, may provide clues to both engines and progenitors. In this paper, we conduct a detailed study of the tight-binary formation scenario for lGRBs, comparing this scenario to other leading progenitor models. Combining this progenitor scenario with different lGRB engines, we can compare to existing data and make predictions for future observational tests. We find that the combination of the tight-binary progenitor scenario with the black hole accretion disk engine can explain lGRBs, low-luminosity GRBs, ultra-long GRBs, and Ic-BL. We discuss the various progenitor properties required for these different subclasses and note such systems would be future gravitational-wave merger sources. We show that the current literature on other progenitor-engine scenarios cannot explain all of these transient classes with a single origin, motivating additional work. We find that the tight-binary progenitor with a magnetar engine is excluded by existing observations. The observations can be used to constrain the properties of stellar evolution, the nature of the GRB, and the associated SN engines in lGRBs and Ic-BL. We discuss the future observations needed to constrain our understanding of these rare, but powerful, explosions.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

DeepClone: Lightweight State Replication of Deep Learning Models for Data Parallel Training

Training modern deep neural network (DNN) models involves complex workflows triggered by model exploration, sensitivity analysis, explainability, etc. A key primitive in this context is the ability to clone a model training instance, i.e. “fork” the training process in a potentially different direction, which enables comparisons of different evolution paths using variations of training data and model parameters. However, in a quest improve the training throughput, a mix of data parallel, model parallel, pipeline parallel and layer-wise parallel approaches are making the problem of cloning highly complex. In this paper, we explore the problem of efficient cloning under such circumstances. To this end, we leverage several properties of data-parallel training and layer-wise parallelism to design DeepClone, a cloning approach based on augmenting the execution graph to gain direct access to tensors, which are then sharded and reconstructed asynchronously in order to minimize runtime overhead, standby duration, readiness duration. Compared with state-of-art approaches, DeepClone shows orders of magnitude improvement for several classes of DNN models.

data-parallel training↗
DOE OSTI2023

Search for a new $Z$' gauge boson in 4$μ$ events with the ATLAS experiment

This paper presents a search for a new Z' vector gauge boson with the ATLAS experiment at the Large Hadron Collider using pp collision data collected at $\sqrt{s}$ = 13 TeV, corresponding to an integrated luminosity of 139 fb -1 . The new gauge boson Z' is predicted by Lμ - Lτ models to address observed phenomena that cannot be explained by the Standard Model. The search examines the four-muon (4μ) final state, using a deep learning neural network classifier to separate the Z' signal from the Standard Model background events. The di-muon invariant masses in the 4μ events are used to extract the Z' resonance signature. No significant excess of events is observed over the predicted background. Upper limits at a 95% confidence level on the Z' production cross-section times the decay branching fraction of pp → Z'μμ → 4μ are set from 0.31 to 4.3 fb for the Z' mass ranging from 5 to 81 GeV. The corresponding common coupling strengths, g Z' , of the Z' boson to the second and third generation leptons above 0.003 – 0.2 have been excluded.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2025

Knowledge Oriented Graph Unified Transformer (KOGUT) v0.1

KOGUT — Knowledge Oriented Graph Unified Transformer KOGUT implements the Relational Graph Transformer (RelGT) architecture for knowledge graph link prediction in biological domains, with a primary focus on microbial growth media prediction. While the original RelGT (arXiv:2505.10960) targets relational tables, time series, and multi-table databases, KOGUT adapts this architecture for heterogeneous biological knowledge graphs, providing first-in-class AI predictive models for microbial cultivation. Key Adaptations Beyond Original RelGT: - Knowledge Graph Focus: Applied to biological KGs with semantic node types (taxa, chemicals, media, phenotypes, environments) versus generic relational database tables, trained on the KG-Microbe knowledge graph (1.3M entities, 2.9M edges, 24 relation types). - Multimodal Node Encoding: Integrates node labels, categories, descriptions, and synonyms from KG metadata through learned embedding layers—adapting relational column features to graph node attributes with textual semantics. - Extended K-Hop Subgraph Strategy: Optimized neighborhood sampling (3-hop default, configurable up to 200 nodes) tuned for sparse biological networks, building on the original local-global attention framework with biological relation preservation. - Biolink Predicate Preservation: Type-specific transformations for 24 biological edge semantics (occurs_in, consumes, produces, has_phenotype, subclass_of) beyond standard relational foreign keys, enabling multi-relation link prediction. - Inductive Learning Support: Enables zero-shot predictions for novel taxa through feature-based embeddings (temperature, oxygen requirements, gram stain, cell shape), extending the original transductive relational benchmark scope to uncultured microorganisms. CheapSOTA Performance Optimizations (This Distribution): - VQ-EMA Centroid Attention: Vector quantization with exponential moving average for improved global context modeling (+5-10% MRR improvement). - HDF5 Precomputed Data Loading: One-time preprocessing of k-hop subgraphs to eliminate redundant graph traversals (2-5× training speedup). - Distributed Data Parallel Training: Multi-GPU support for scaling to larger knowledge graphs (tested on 4× NVIDIA A100 GPUs at NERSC Perlmutter). - Mixed Precision Training: Automatic mixed precision (AMP) for memory efficiency and faster training. Advantages Over Standard Knowledge Graph Embedding Models: Combines RelGT's proven multi-element tokenization (features, type, hop, structure) with graph-native biological representations, enabling interpretable link prediction across heterogeneous entities that standard embedding models (TransE, RotatE, ComplEx) and table-based transformers cannot directly model. Achieves near-perfect performance on microbial growth media prediction (MRR: 0.9966, Precision@1: 0.9932, Hit@10: 1.0000) while maintaining explainability through attention-based reasoning over biological pathways. Training Data: - KG-Microbe merged knowledge graph: 1,379,337 nodes, 2,960,472 edges - 24 biological relation types including taxonomic hierarchies, metabolic interactions, phenotype associations, and environmental relationships - Primary prediction task: Growth media suitability for microbial taxa (biolink:occurs_in, 50K edges) - Multi-relation capability: Predicts links for any of the 24 relation types, including chemical consumption/production, phenotype associations, and taxonomic classification Citation: Original RelGT Architecture: Dwivedi et al., "Relational Graph Transformer", arXiv:2505.10960, 2025 KOGUT Implementation: Knowledge Oriented Graph Unified Transformer for Microbial Growth Media Prediction Developed at Lawrence Berkeley National Laboratory (LBNL) Trained on NERSC Perlmutter supercomputer

Joachimiak, Marcin [Lawrence Berkeley National Lab↗
DOE OSTI2025

Public Reference Data for Megawatt-Scale Hydrogen Electrolysis – Simulated Wind

The U.S. Department of Energy and the National Laboratory of the Rockies (NLR) demonstrate hydrogen electrolysis, hydrogen compression and storage, and variable hydrogen fuel cell power production using megawatt-scale equipment at NLR’s Flatirons Campus as part of the Advanced Research on Integrated Energy Systems (ARIES) initiative. This dataset represents part of that effort and is intended for academic, national laboratory, industrial, and other stakeholders to plan, design, and validate models of megawatt-scale hydrogen technologies and diverse energy infrastructure nationwide. These data provide a baseline for how existing hydrogen electrolysis technologies perform when coupled with various energy technologies. Future datasets will demonstrate how existing hydrogen fuel cell technologies can provide controllable, dispatchable, and variable power output for artificial intelligence (AI) data centers and other variable loads. This dataset entry describes hydrogen production using a single, simulated wind turbine. The electrolyzer is a 1.25-MW proton exchange membrane type MC250 system manufactured by Nel Hydrogen . While the unit supports up to 2.5 MW of electrolysis, NLR only has a single 1.25-MW electrolysis stack. For the simulated wind energy profiles, NLR used OpenFAST to simulate a 3.4-MW International Energy Agency (IEA) reference wind turbine. The hour-long wind energy profiles varied over wind turbulence intensity (Class A or Class C) and average wind speed (5, 7, or 9 m/s). To match the power limits of the 1.25-MW electrolyzer and 3.4-MW IEA wind turbine most effectively and to maximize the efficiency of hydrogen production at a given average wind speed, the profiles were sometimes scaled by two times. This means that, in some cases, the experimental setup assumed two 1.25-MW electrolyzers were coupled with the wind turbine, representing a total maximum electrolysis load of 2.5 MW. Finally, NLR experimented with two settings for the electrolyzer power supply minimum and maximum current ramp rates (gain and slew): 200 and 400 amperes per second. The simulated profiles were translated from power (kilowatts) to current (amperes) using a curve fit with calibration data and sent to the electrolyzer power supply at 1-Hz frequency. These datasets report relevant hydrogen balance-of-plant and system data, all captured at 1 Hz, including hydrogen mass production measured with an Emerson Coriolis flow meter. Each .zip file represents a single wind turbine electrolysis experiment and is formatted as follows: {technology}-{average wind speed}-{turbulence class}_{number of 1.25 MW electrolyzers connected}-{electrolyzer ramp rate in amperes/second} For instance, “windIEA3.4-5ms-C_2-400.zip” represents the hour-long experiment using the IEA 3.4-MW turbine, subjected to an average wind speed of 5 m/s and Class C wind turbulence, and connected to two 1.25-MW electrolyzers with the power supply set to a maximum current ramp rate (gain and slew) of 400 A/s. Each .zip folder contains the following files: A .csv file containing raw data. An .xlsx file explaining all the fields in the raw data. A .png plot showing the time series of hydrogen production in kilograms per hour, electrolysis power consumption, and input wind turbine power. An experiment labeled “characterization_200.zip” demonstrates the MC250 electrolyzer steady-state response with 30 minute load steps for a total duration of 5 hours. Finally, a .csv file is provided with all simulated wind experiments combined into one dataset labeled "combined_wind_experiments.csv". NLR also built an AI/machine-learning predictive model based on these datasets. The model ingests the electrolyzer current command in amperes, as well as various pressures and temperatures across the system, and predicts hydrogen output in kilograms per hour. The complete model can be found at https://huggingface.co/NatLabRockies/ptmelt-hydrogen-electrolysis .

08 HYDROGEN↗
DOE OSTI2025

Public Reference Data for Megawatt-Scale Hydrogen Electrolysis - Simulated Wave

The U.S. Department of Energy and the National Laboratory of the Rockies (NLR) demonstrate hydrogen electrolysis, hydrogen compression and storage, and variable hydrogen fuel cell power production using megawatt-scale equipment at NLR’s Flatirons Campus as part of the Advanced Research on Integrated Energy Systems (ARIES) initiative. This dataset represents part of that effort and is intended for academic, national laboratory, industrial, and other stakeholders to plan, design, and validate models of megawatt-scale hydrogen technologies and diverse energy infrastructure nationwide. These data provide a baseline for how existing hydrogen electrolysis technologies perform when coupled with various energy technologies. Future datasets will demonstrate how existing hydrogen fuel cell technologies can provide controllable, dispatchable, and variable power output for artificial intelligence (AI) data centers and other variable loads. This dataset entry describes hydrogen production using a single, simulated wave energy conversion device. The electrolyzer is a 1.25-MW proton exchange membrane type MC250 system manufactured by Nel Hydrogen. While the unit supports up to 2.5 MW of electrolysis, NLR only has a single 1.25-MW electrolysis stack. For the wave energy, NLR used a wave energy converter model from PacWave. These devices can be equipped with accumulators and pressure relief values to smooth the power output by storing and releasing hydraulic energy. Using a peak power output of 10 MW, the model created two 25-minute profiles: one with and one without the accumulators and pressure relief valves. To down select the profile data from the native resolution of 20 Hz to 1 Hz, NLR took the mean of every 20 data points. NLR experimented with two simulated wave energy power plants: one that peaks at 10 MW, and one that peaks at 5 MW. These profiles were scaled for the physical 1.25 MW electrolyzer by multiplying the original profiles by one eighth and one quarter, respectively. The first profile matches the capacity rating of eight of the 1.25 MW electrolyzers, while the second matches four electrolyzers. Finally, NLR experimented with two settings for the electrolyzer power supply minimum and maximum current ramp rates (gain and slew): 200 and 400 amperes per second. The simulated profiles were translated from power (kilowatts) to current (amperes) using a curve fit with calibration data and sent to the electrolyzer power supply at 1-Hz frequency. These datasets report relevant hydrogen balance-of-plant and system data, all captured at 1 Hz, including hydrogen mass production measured with an Emerson Coriolis flow meter. Each .zip file represents a single wave electrolysis experiment and is formatted as follows: {technology}-{accumulator?}_{number of 1.25 MW electrolyzers connected}-{electrolyzer ramp rate in amperes/second} For instance, “wavePacWave-Noacc_4-400.zip” represents the 25 minute-long experiment using the PacWave’s wave energy converter model, equipped with no accumulator, connected to four 1.25-MW electrolyzers with their power supplies set to a maximum current ramp rate (gain and slew) of 400 A/s. Each .zip folder contains the following files: A .csv file containing raw data. An .xlsx file explaining all the fields in the raw data. A .png plot showing the time series of hydrogen production in kilograms per hour, electrolysis power consumption, and input wave power. An experiment, labeled “characterization_200.zip”, demonstrates the MC250 electrolyzer steady-state response with 30 minute load steps for a total duration of 5 hours. Finally, a .csv file is provided with all wave profiles combined into one dataset labeled "combined_wave_experiments.csv". NLR also built an AI/machine-learning predictive model based on these datasets. The model ingests the electrolyzer current command in amperes, as well as various pressures and temperatures across the system, and predicts hydrogen output in kilograms per hour. The complete model can be found at https://huggingface.co/NatLabRockies/ptmelt-hydrogen-electrolysis.

08 HYDROGEN↗