Search NASA⌕ Search

SEARCH · Search NASA

Results for “modeling discrepancy”

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 91 records · Page 5
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

An Assessment of Kinetic Models for Ammonia Flame Extinction

The growing interest in using ammonia as a carbon-neutral replacement fuel has prompted numerous recent research efforts towards understanding fundamental combustion characteristics and chemical kinetics of ammonia oxidation. Despite these recent efforts, there remains a large deficiency in extinction limit data for ammonia flames, particularly under elevated pressure and temperature conditions relevant to gas turbines. Even the literature reaction mechanisms developed specifically for ammonia have been untested under gas turbine relevant conditions, and existing kinetic models have shown large variations in laminar flame speed, extinction strain rate, and speciation predictions for NH3–H2 mixtures at ambient conditions. Recent work by Thomas and co-workers, for example, compared different ammonia kinetic models in the literature for predicting their extinction strain rate measurements for non-premixed NH3-H2 counterflow flames, and again have highlighted the discrepancies in model predictions. Here, our work has undertaken to further highlight important reaction pathways for ammonia oxidation that may be responsible for these discrepancies for extinction strain rate predictions, as well as elucidate sensitive reaction steps with large uncertainties that require further attention. This work points to a need for an improved understanding of ammonia combustion chemistry as well as new datasets at elevated pressure and temperature conditions to better constrain rate expressions for key reaction steps.

ammonia, flame, combustion, kinetic, kinetics↗
DOE OSTI2023

Reproducible benchmark for the SNAP 8 experimental reactor at dry conditions

Here, this work provides fully reproducible benchmark models of the Systems for Nuclear Auxiliary Power (SNAP) 8 Experimental Reactor. Validation benchmarks of the criticality configuration experiments under dry conditions with no coolant are presented. In addition, relevant reactivity worth experiments, and measurement of power distributions are validated and presented here. Discrepancies between modeled and experimental results are at most 110 pcm for fuel and poison worths, and critical configurations that do not manipulate control elements are within 50 pcm. Larger discrepancies found in reflector element worth experiments are due to experimental methodology which was noted as being simplified due to limited calculational capabilities at the time, while discrepancy in the control element manipulated critical configuration is due the absence of structural components for modeling simplification. Power distributions closely follow what was observed in experiment with expected peaking in reflecting elements. All models are well documented with cited references describing and justifying assumptions, simplifications, and adjustments to reproduce the models with any nuclear safety codes; model inputs and outputs are stored in the snapReactors GitHub repository.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2021

Understanding water and energy fluxes in the Amazonia: Lessons from an observation‐model intercomparison

Abstract Tropical forests are an important part of global water and energy cycles, but the mechanisms that drive seasonality of their land‐atmosphere exchanges have proven challenging to capture in models. Here, we (1) report the seasonality of fluxes of latent heat (LE), sensible heat ( H ), and outgoing short and longwave radiation at four diverse tropical forest sites across Amazonia—along the equator from the Caxiuanã and Tapajós National Forests in the eastern Amazon to a forest near Manaus, and from the equatorial zone to the southern forest in Reserva Jaru; (2) investigate how vegetation and climate influence these fluxes; and (3) evaluate land surface model performance by comparing simulations to observations. We found that previously identified failure of models to capture observed dry‐season increases in evapotranspiration (ET) was associated with model overestimations of (1) magnitude and seasonality of Bowen ratios (relative to aseasonal observations in which sensible was only 20%–30% of the latent heat flux) indicating model exaggerated water limitation, (2) canopy emissivity and reflectance (albedo was only 10%–15% of incoming solar radiation, compared to 0.15%–0.22% simulated), and (3) vegetation temperatures (due to underestimation of dry‐season ET and associated cooling). These partially compensating model‐observation discrepancies (e.g., higher temperatures expected from excess Bowen ratios were partially ameliorated by brighter leaves and more interception/evaporation) significantly biased seasonal model estimates of net radiation ( R n ), the key driver of water and energy fluxes (LE ~ 0.6 R n and H ~ 0.15 R n ), though these biases varied among sites and models. A better representation of energy‐related parameters associated with dynamic phenology (e.g., leaf optical properties, canopy interception, and skin temperature) could improve simulations and benchmarking of current vegetation–atmosphere exchange and reduce uncertainty of regional and global biogeochemical models.

Restrepo‐Coupe, Natalia↗
DOE OSTI2023

Thermal transport in warm dense matter revealed by refraction-enhanced x-ray radiography with a deep-neural-network analysis

Abstract Transport properties of high energy density matter affect the evolution of many systems, ranging from the geodynamo in the Earth’s core, to hydrodynamic instability growth in inertial confinement fusion capsules. Large uncertainties of these properties are present in the warm dense matter regime where both plasma models and condensed matter models become invalid. To overcome this limit, we devise an experimental platform based on x-ray differential heating and time-resolved refraction-enhanced radiography coupled to a deep neural network. We retrieve the first measurement of thermal conductivity of CH and Be in the warm dense matter regime and compare our measurement with the most commonly adopted models. The discrepancies observed are related to the estimation of a correction term from electron-electron collisions. The results necessitate improvement of transport models in the warm dense matter regime and could impact the understanding of the implosion performance for inertial confinement fusion.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2023

P2M Simulation Exercise on Past Fuel Melting Irradiation Experiments

This paper presents the results of the Power To Melt and Maneuverability (P2M) Simulation Exercise on past fuel melting irradiation experiments, organized within the Organisation for Economic Co-operation and Development/Nuclear Energy Agency Framework for IrraDiation ExperimentS (FIDES) framework by the Core Group (CEA, EDF, and SCK·CEN) and open to all FIDES members. The exercise consisted in simulating two past power ramps where fuel melting was detected: (1) the xM3 staircase power transient [ramp terminal level (RTL) 70 kW·m -1 , average burnup 27 GWd·tU -1 ], carried out in 2005 in the R2 reactor at Studsvik (Sweden), where the rodlet maintained its integrity, and (2) the HBC4 fast power transient (RTL 66 kW·m -1 , average burnup 48 GWd·tU -1 ), carried out in 1987 in the BR2 reactor at SCK·CEN (Belgium), where the cladding failed during the experiment. The exercise was joined by 13 organizations from 9 countries using 11 different fuel performance codes. In this paper, the main results of the Simulation Exercise are presented and compared to available postirradiation examinations (PIE) or on-line measurements during the power ramps (fuel and clad diameters, rod elongation, pellet-clad gap, and fission gas release). Since the focus of the Simulation Exercise is on fuel melting assessment, determination of the boundary between melted/nonmelted fuel and the consequent definition of a melting radius from PIE are first discussed. During the HBC4 ramp, fuel melting was predicted by most of the codes despite differences in the melting models. Higher discrepancies were observed for the xM3 rod that can be attributed partly to power uncertainty and partly to the limited capability of the models to describe partial melting of the fuel during this ramp. Finally, possible code developments to improve simulation results are presented.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2021

Application of Machine Learning to Predict the Response of the Liquid Mercury Target at the Spallation Neutron Source

The Spallation Neutron Source (SNS) at Oak Ridge National Laboratory is currently the most powerful accelerator-driven neutron source in the world. The intense proton pulses strike on SNS’s mercury target to provide bright neutron beams, which also leads to severe fluid-structure interactions inside the target. Prediction of resultant loading on the target is difficult particularly when helium gas is intentionally injected into mercury to reduce the loading and mitigate the pitting damage on the target’s internal walls. Leveraging the power of machine learning and the measured target strain, we have developed machine learning surrogates for modeling the discrepancy between simulations and experimental strain data. We then employ these surrogates to guide the refinement of the high-fidelity mercury/helium mixture model to predict a better match of target strain response.

Lin, Lianshan↗
DOE OSTI2025

Experiment-modeling studies comparing energy dissipation in the DIII-D SAS and SAS-VW divertors

Recent DIII-D experiments on Small Angle Slot (SAS) divertors have confirmed that a combination of divertor closure and target shaping can enhance cooling across the divertor target and increase energy dissipation, but with significant dependence on B T (toroidal magnetic field) direction. In these novel divertors, the roles of closure, target shaping, drifts, and scale lengths are all interconnected in optimizing dissipation, with the separatrix electron density n eSEP being the key parameter associated with the level of dissipation/detachment. After modifying the original flat-targeted graphite SAS to include a V shape with a tungsten coating on the outer side of the divertor (SAS-VW), matched series of discharges were run to compare to detailed SOLPS-ITER modeling. Experimentally, when run as designed with the outer strike point at the slot vertex, SAS-VW requires nearly identical n eSEP for detachment as the original SAS, with little difference in dissipation for the new geometry. This is in contrast to (1) earlier modeling predictions that a small change of the SAS geometry to a V shape should enhance dissipation at the same n eSEP for magnetic configurations having better H-mode access (ion B × ∇B drift directed into the divertor), and (2) despite the achievement of significantly higher (2-7x) neutral pressures and compression in the SAS-VW slot. Comparisons of experimental density scans to the most recent SOLPS-ITER modeling with ExB drifts show reasonable agreement for dissipation/detachment onset when using separatrix density as the independent parameter. In order to help understand the discrepancy in modeled vs actual performance for the new configuration, additional measurements varying gas injection location and impurity injection were undertaken. In-slot D 2 gas fueling is more effective (5–22 %) in promoting detachment, in accord with modeling. In-slot impurity injection (N 2 or Ne) can yield 30 % lower core Z eff and 15 % less confinement degradation after detachment compared to main chamber puffing, as well as relatively lower tungsten leakage from the divertor. Modeling can also reproduce the improved detachment seen as the strike point moves inboard of the slot vertex. While we can explain the effects of the most important parameters causing energy dissipation in these slot divertors, it remains that many aspects of their behavior cannot be accurately modeled using state-of-art codes such as SOLPS-ITER. This is of concern for future model-driven designs utilizing similar V-shaped geometries.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2025

Modeling 2p-2h Interactions in Neutrino-Nucleus Scattering: A Comparative Overview of the Valencia and SuSAv2-MEC Models

2p-2h interactions are crucial for describing ν-A scattering, especially in the "dip region" between the Quasi-Elastic (QE) peak and Δ-resonance production. They involve the ejection of two nucleons, leading to two holes in the nuclear ground state. The underlying nuclear dynamics involve short-range correlations (SRC), long-range correlations (Random Phase Approximation, RPA), and the interplay of one- and two-body currents. Several models attempt to describe 2p-2h interactions, including Microscopic Models (e.g., Valencia), Scaling-Based Models (e.g., SuSAv2-MEC), and Transport Models (e.g., GiBUU). Despite progress, discrepancies exist between models and with data. I Will give an overview of 2p2h from the perspective of two Models Valencia and SuSAv2-MEC.

Hassinin, Karim [Houston U.]↗
DOE OSTI2025

A Model Intercomparison Study of Aerosol‐Cloud‐Turbulence Interactions in a Cloud Chamber: 1. Model Results

This study presents the first model intercomparison of aerosol‐cloud‐turbulence interactions in a controlled cloudy Rayleigh‐Bénard Convection chamber environment, utilizing the Pi Chamber at Michigan Technological University. We analyzed simulated cloud chamber‐averaged statistics of microphysics and thermodynamics in a warm‐phase, cloudy environment under steady‐state conditions at varying aerosol injection rates. Simulation results from seven distinct models (DNS, LES, and a 1D turbulence model) were compared. Our findings demonstrate that while all models qualitatively capture observed trends in droplet number concentration, mean radius, and droplet size distributions at both high and low aerosol injection rates, significant quantitative differences were observed. Notably, droplet number concentrations varied by over two orders of magnitude between models for the same injection rates, indicating sensitivities to the model treatments in droplet activation and removal and wall fluxes. Furthermore, inconsistencies in vertical relative humidity profiles and in achieving steady‐state liquid water content suggest the need for further investigation into the mechanisms driving these variations. Despite these discrepancies, the models generally reproduced consistent power‐law relationships between the microphysical variables. This model intercomparison underscores the importance of controlled cloud chamber experiments for validating and improving cloud microphysical parameterizations. Recommendations for future modeling studies are also highlighted, including constraining wall conditions and processes, investigating droplet/aerosol removal (including sidewall losses), and conducting simplified experiments to isolate specific processes contributing to model divergence and reduce model uncertainties.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Basic Research Needs for Inverse Methods for Complex Systems under Uncertainty

Inverse problems, which aim to infer unknown properties of a system using experimental and observational data, are central to addressing many of the U.S. Department of Energy’s (DOE) most critical scientific and engineering challenges. Accurate, computationally efficient, and data-efficient solutions to inverse problems are essential for advancing DOE mission-critical science drivers, including analyzing data from large-scale experimental facilities, optimizing fusion reactor performance, accelerating materials discovery, enhancing geophysical imaging, improving wildfire predictions, and enabling autonomous systems and digital twins. However, these problems are becoming increasingly complex, often involving nonlinear, highdimensional, and interconnected systems and models that span multiple physics and scales, while relying on data with varying quantity, quality, and information content. Compounding these challenges is the uncertainty inherent in DOE-relevant systems, where errors in inputs, noise in data, incompleteness of data, and discrepancies between models and reality constrain the accuracy and precision of solutions. At the same time, the convergence of recent scientific computing trends—scientific machine learning, artificial intelligence, and computing advances such as exascale computing—is creating unprecedented opportunities for tackling these challenges. The cross-cutting nature of inverse problems, combined with their growing complexity and rapidly evolving data and algorithmic demands, strongly motivates the formulation of a prioritized research agenda to maximize their capabilities and impact. In response to this need, DOE’s Advanced Scientific Computing Research (ASCR) program in the Office of Science convened the Workshop on Basic Research Needs for Inverse Problems for Complex Systems Under Uncertainty in June 2025. This workshop brought together experts across disciplines to identify grand challenges and major opportunities in the field. Through collaborative discussions, the workshop defined transformative research directions aimed at addressing the mathematical, statistical, and computational challenges posed by inverse problems under uncertainty. As a result of these efforts, four priority research directions (PRDs) were identified to guide future research and development in this area. These PRDs, summarized below, represent a roadmap for advancing the foundational science and mathematics of inverse problems, enabling robust, scalable, and uncertainty-aware solutions that are critical for DOE applications.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

22 Ne Phase Separation as a Solution to the Ultramassive White Dwarf Cooling Anomaly

The precise astrometric measurements of the Gaia Data Release 2 have opened the door to detailed tests of the predictions of white dwarf cooling models. Significant discrepancies between theory and observations have been identified, the most striking affecting ultramassive white dwarfs. Cheng et al. found that a small fraction of white dwarfs on the so-called Q branch must experience an extra cooling delay of ~8 Gyr not predicted by current models. 22 Ne phase separation in a crystallizing C/O white dwarf can lead to a distillation process that efficiently transports 22 Ne toward its center, thereby releasing a considerable amount of gravitational energy. Here, using state-of-the-art Monte Carlo simulations, we show that this mechanism can largely resolve the ultramassive cooling anomaly if the delayed population consists of white dwarfs with moderately above-average 22 Ne abundances. We also argue that 22 Ne phase separation can account for the smaller cooling delay currently missing for models of white dwarfs with more standard compositions.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

Machine learning based algorithms for uncertainty quantification in numerical weather prediction models

Complex numerical weather prediction models incorporate a variety of physical processes, each described by multiple alternative physical schemes with specific parameters. The selection of the physical schemes and the choice of the corresponding physical parameters during model configuration can significantly impact the accuracy of model forecasts. There is no combination of physical schemes that works best for all times, at all locations, and under all conditions. It is therefore of considerable interest to understand the interplay between the choice of physics and the accuracy of the resulting forecasts under different conditions. This paper demonstrates the use of machine learning techniques to study the uncertainty in numerical weather prediction models due to the interaction of multiple physical processes. The first problem addressed herein is the estimation of systematic model errors in output quantities of interest at future times, and the use of this information to improve the model forecasts. The second problem considered is the identification of those specific physical processes that contribute most to the forecast uncertainty in the quantity of interest under specified meteorological conditions. In order to address these questions we employ two machine learning approaches, random forests and artificial neural networks. The discrepancies between model results and observations at past times are used to learn the relationships between the choice of physical processes and the resulting forecast errors. Numerical experiments are carried out with the Weather Research and Forecasting (WRF) model. The output quantity of interest is the model precipitation, a variable that is both extremely important and very challenging to forecast. The physical processes under consideration include various micro-physics schemes, cumulus parameterizations, short wave, and long wave radiation schemes. The experiments demonstrate the strong potential of machine learning approaches to aid the study of model errors.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

The Role of Internal Variability in Springtime Arctic Amplification from 1980 to 2022

Arctic amplification (AA) refers to the enhanced warming of the Arctic relative to the global average due to rising greenhouse gases, measured as the ratio of Arctic-mean to global-mean surface air temperature (SAT) trends. From 1980 to 2022, annual-mean AA reached 4.2 (Arctic defined as north of 70°N). Climate models simulate AA but fail to reproduce its magnitude. Sweeney et al. attributed much of this model–observation discrepancy to internal variability. AA shows seasonality and so does the discrepancy. Spring (March–May) shows the largest gap: Observed AA is 4.2, while the multimodel mean is 2.7. This raises several questions: 1) What role does internal variability play in observed spring AA? 2) How does simulated spring AA compare to observations when internal variability is removed? 3) If internal variability is significant, what mechanisms drive it? To address these, we adapted the machine learning algorithm from Sweeney et al., training on simulated multidecadal spring SAT and sea level pressure (SLP) trend maps. Our results show that internal variability enhanced spring Arctic warming by 37% and reduced global warming by 10%. Removing internal variability reconciles the spring AA discrepancy. The estimated internal contribution to Arctic spring warming is supported by an independent dynamical adjustment approach. We identify an atmospheric circulation pattern in observations associated with this internal warming. Observed internal Siberian SAT and SLP trends follow the simulated SAT–SLP relationship but lie at the distribution’s extreme, suggesting models generally underestimate internal variability unless the observed configuration reflects a rare real-world realization.

Arctic↗
DOE OSTI2023

Bi-fidelity modeling of uncertain and partially unknown systems using DeepONets

Recent advances in modeling large-scale, complex physical systems have shifted research focuses towards data-driven techniques. However, generating datasets by simulating complex systems can require significant computational resources. Similarly, acquiring experimental datasets can prove difficult. For these systems, often computationally inexpensive, but in general inaccurate models, known as the low-fidelity models, are available. Here in this paper, we propose a bi-fidelity modeling approach for complex physical systems, where we model the discrepancy between the true system's response and a low-fidelity response in the presence of a small training dataset from the true system's response using a deep operator network, a neural network architecture suitable for approximating nonlinear operators. We apply the approach to systems that have parametric uncertainty and are partially unknown. Three numerical examples are used to show the efficacy of the proposed approach to model uncertain and partially unknown physical systems.

MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Toward engineering lattice structures with the material point method (MPM)

This study examines the potential of two variants of the material point method—the generalized interpolation material point (GIMP) and dual domain material point (DDMP) methods—in developing a robust computational framework for engineering lattice structures under different loading conditions. The study begins with assessing the ability of the two methods in predicting elastic buckling phenomena using column geometries with and without initial geometric imperfections. The results indicate that both methods effectively capture buckling phenomena when initial geometric imperfections are introduced. After this verification step, we create several models of tetrahedral lattice structures with varying strut diameter and orientation and subject them to quasi-static loading. We then validate the numerical results using laboratory test results. The results show that, while both methods accurately predict load–displacement curves in the pre-buckling regime, their predictive capabilities diminish in the post-buckling regime. Through visual comparison between the numerical and experimental deformed shapes, it appears that the discrepancies between model and experimental results are attributed to initial geometric imperfections in the lattices that occurred during 3D printing. We then establish a second set of lattice models where different types of initial geometric imperfections are considered. The results from these models show that imperfections have a negligible influence in the pre-buckling regime but affect the behavior considerably in the post-buckling regime. As a final step in this work, we subject the lattice models to impact loading and employ hypothetical soft and stiff materials. These results show that the lattice stiffness, which depends on material stiffness, strut diameter, and orientation, significantly influences the ability of a lattice structure to resist impact. In particular, we find that a stiffer lattice (i.e., one made with a stiff material and thicker struts) is capable of absorbing more energy than a softer one during impact. Although material nonlinearities, inelasticity, and detailed contact formulations are not considered in this study, the findings obtained herein lay the groundwork for engineering lattice structures under extreme loading conditions through a simulation-driven framework based on particle-based methods.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Conformer-Dependent Chemistry: Experimental Product Branching of the Vinyl Alcohol + OH + O 2 Reaction

The concentration of formic acid in Earth's tropo-sphere is underestimated by detailed chemical models compared to field observations. Phototautomerization of acetaldehyde to its less stable tautomer vinyl alcohol, followed by the OH-initiated oxidation of vinyl alcohol, has been proposed as a missing source of formic acid that improves the agreement between models and field measurements. Theoretical investigations of the OH + vinyl alcohol reaction in excess O 2 conclude that OH addition to the alpha carbon of vinyl alcohol produces formaldehyde + formic acid + OH, whereas OH addition to the β site leads to glycoaldehyde + HO 2 . Furthermore, these studies predict that the conformeric structure of vinyl alcohol controls the reaction pathway, with the anti-conformer of vinyl alcohol promoting α OH addition, whereas the syn-conformer promotes β addition. However, the two theoretical studies reach different conclusions regarding which set of products dominate. We studied this reaction using time-resolved multiplexed photoionization mass spectrometry to quantify the product branching fractions. Our results, supported by a detailed kinetic model, conclude that the glycoaldehyde product channel (arising mostly from syn-vinyl alcohol) dominates over formic acid production with a 3.6:1.0 branching ratio. Additionally, this result supports the conclusion of Lei et al. that conformer-dependent hydrogen bonding at the transition state for OH-addition controls the reaction outcome. As a result, tropospheric oxidation of vinyl alcohol creates less formic acid than recently thought, increasing again the discrepancy between models and field observations of Earth's formic acid budget.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Machine learning for detection of 3D features using sparse x-ray tomographic reconstruction

In many inertial confinement fusion (ICF) experiments, the neutron yield and other parameters cannot be completely accounted for with one and two dimensional models. This discrepancy suggests that there are three dimensional effects that may be significant. Sources of these effects include defects in the shells and defects in shell interfaces, the fill tube of the capsule, and the joint feature in double shell targets. Due to their ability to penetrate materials, x rays are used to capture the internal structure of objects. Methods such as computational tomography use x-ray radiographs from hundreds of projections, in order to reconstruct a three dimensional model of the object. In experimental environments, such as the National Ignition Facility and Omega-60, the availability of these views is scarce, and in many cases only consists of a single line of sight. Mathematical reconstruction of a 3D object from sparse views is an ill-posed inverse problem. These types of problems are typically solved by utilizing prior information. Neural networks have been used for the task of 3D reconstruction as they are capable of encoding and leveraging this prior information. We utilize half a dozen, different convolutional neural networks to produce different 3D representations of ICF implosions from the experimental data. Deep supervision is utilized to train a neural network to produce high-resolution reconstructions. These representations are used to track 3D features of the capsules, such as the ablator, inner shell, and the joint between shell hemispheres. Machine learning, supplemented by different priors, is a promising method for 3D reconstructions in ICF and x-ray radiography, in general.

Wolfe, Bradley T. (ORCID:0000000268301614)↗
DOE OSTI2022

Kinetic model development and Bayesian uncertainty quantification for the complete reduction of Fe-based oxygen carriers with CH 4 , CO, and H 2 for chemical looping combustion

In this work, three kinetic models are developed and calibrated for the complete multi-step reduction of an Fe-based oxygen carrier (OC) particle with CH 4 , CO, and H 2 , using data from thermogravimetric analysis. The complete reduction rate profiles exhibit complex dynamics whose trajectory is significantly different depending on the reducing gas. A Bayesian model building and parameter estimation framework is applied for simultaneous parameter and model structure uncertainty quantification. The final models show excellent agreement between model predictions and calibration data, as well as new data not used for calibration (for the reduction of the OC with CH 4 ). Parameter uncertainty is quantified by determining joint posterior distribution, and model structure uncertainty is addressed by incorporating Gaussian process stochastic functions (represented by Bayesian smoothing splines) into the kinetic models. The final kinetic models with discrepancy functions are readily employable in equation-oriented simulation and optimization platforms.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗