Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “heterogeneous computing”

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 685 records · Page 38
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

A Performance-Portable MultiGPU Implementation of 3D Euler Equations using ProtoX and IRIS

Computational scientists often face challenges when developing and optimizing code for high-performance computing (HPC), especially when trying to leverage GPUs. Given the heterogeneity of the nodes that comprise many modern HPC facilities, considerable demand exists for performance portable solutions for the core computational kernels used in many scientific computing libraries. In this work, we demonstrate a fourth-order finite volume method–based implementation of the Euler equations, which are an integral part of computational fluid dynamics. Our performance-portable multiGPU implementation for Euler equations uses ProtoX to generate kernels and IRIS for portability. ProtoX is a domain-specific language that uses a structured-grid partial differential equation library called Proto as its front end and the SPIRAL code generation system as its back end to generate optimized kernels for different architectures. Optimized kernels generated by ProtoX are orchestrated through the IRIS intelligent runtime system to provide portability. Two levels of optimizations within the IRIS runtime— directed acyclic graph fusion and task fusion—are explored to efficiently utilize computing resources in a multiGPU environment. Performance improvement through these optimizations is showcased by comparing the base ProtoX-IRIS implementation on AMD GPUs (Frontier node) and on NVIDIA GPUs (NVIDIA DGX-1).

Mankad, Het↗
DOE OSTI2012

Analytical effective tensor for flow-through composites

A machine, method and computer-usable medium for modeling an average flow of a substance through a composite material. Such a modeling includes an analytical calculation of an effective tensor K.sup.a suitable for use with a variety of media. The analytical calculation corresponds to an approximation to the tensor K, and follows by first computing the diagonal values, and then identifying symmetries of the heterogeneity distribution. Additional calculations include determining the center of mass of the heterogeneous cell and its angle according to a defined Cartesian system, and utilizing this angle into a rotation formula to compute the off-diagonal values and determining its sign.

42 ENGINEERING↗
DOE OSTI2017

SharP

The pre-exascale systems are expected to have a significant amount of hierarchical and heterogeneous on-node memory, and this trend of system architecture in extreme-scale systems is expected to continue into the exascale era. along with hierarchical-heterogeneous memory, the system typically has a high-performing network ad a compute accelerator. This system architecture is not only effective for running traditional High Performance Computing (HPC) applications (Big-Compute), but also for running data-intensive HPC applications and Big-Data applications. As a consequence, there is a growing desire to have a single system serve the needs of both Big-Compute and Big-Data applications. Though the system architecture supports the convergence of the Big-Compute and Big-Data, the programming models and software layer have yet to evolve to support either hierarchical-heterogeneous memory systems or the convergence. A programming abstraction to address this problem. The programming abstraction is implemented as a software library and runs on pre-exascale and exascale systems supporting current and emerging system architecture. Using distributed data-structures as a central concept, it provides (1) a simple, usable, and portable abstraction for hierarchical-heterogeneous memory and (2) a unified programming abstraction for Big-Compute and Big-Data applications.

Venkata, Manjunath Gorentla↗
DOE OSTI2023

Gas-Liquid Flow Modeling for Renewable Fuels Production

Aerobic/anaerobic and gas fermentation pathways have emerged as promising new technologies for the generation of renewable fuels/chemicals from biomass derived sugars, and mixtures of greenhouse/energy rich gas streams (CO2/CH4/H2/CO) via microbial action. Example pathways include sugars-to-ethanol conversion, biomethanation (CO2/H2 to CH4), biogas upgrading, CO fermentation and wet-waste conversion. Gas and liquid phase transport, mass-transfer, and mixing physics at large length scales can significantly affect microbial conversion rates, particularly when the microbial reaction requires a narrow set of conditions. These phenomena are difficult to study in small-scale bench-top reactors that are typically well-mixed. Predictive computational fluid dynamics (CFD) based simulations can therefore aid in the scale-up, design and optimization of these reactors. This work presents multiphase Euler-Euler CFD simulations of at-scale (~500 m3) bioreactors. Our mathematical model treats the gas and liquid as interpenetrating phases. This approach reduces the computational complexity of tracking individual gas bubbles that are several orders of magnitude smaller than reactor dimensions. We solve the Reynolds averaged Navier-Stokes (RANS) multiphase equations that account for phase and chemical species transport, interphase mass and momentum transfer and uses a phenomenological model for gas uptake by microbes. We use a customized solver derived from open-source CFD toolbox, OpenFOAM [1], to perform these simulations, which has been validated against small-scale reactors in our previous work [2]. There is currently a knowledge-gap regarding bubble-size distributions when using gas mixtures with vastly different properties, which can have a significant impact overall mass-transfer. For example, hydrogen bubbles are more buoyant compared to other relatively heavier gases (CO2/CH4/CO), resulting in a large distribution of residence times and bubble sizes. This work therefore develops a deeper understanding of bubble dynamics and interphase mass transfer in such heterogenous gas mixtures through well-resolved computational models. We use a population balance model (PBM) for bubble-size-distribution modeling that is validated against small-scale experiments in our solver with an uncertainty quantification study for bubble coalescence and break-up model parameters. Results pertaining to multiple simulations of gas-fermentation reactors are presented where gas mixtures with varying compositions of CO2/CH4/CO/H2 are imposed at the sparger boundaries. The spatio-temporal variations in bubble-size distribution and mass transfer coefficient are analyzed for varying superficial velocities and gas-compositions for varying sizes of bubble-column and airlift reactors. This work will also examine the performance of different reactor designs, viz. bubble column reactor, airlift reactor with an internal draft tube, and a stirred-tank reactor with Rushton impellers. Reactor mass-transfer coefficient, gas hold-up, and dissolved gas distribution are critically analyzed among reactors, and sensitivity studies pertaining to gas flow rates and reactor geometry will be presented. [1] Weller, H., Tabor, G., Jasak, H. and Fureby, C., A tensorial approach to computational continuum mechanics using object-oriented techniques, Computers in physics, 12, 6, 620--631, 1998. [2] Rahimi, M., Sitaraman, H., Humbird, D. and Stickel, J., Computational fluid dynamics study of full-scale aerobic bioreactors: Evaluation of gas-liquid mass transfer, oxygen uptake, and dynamic oxygen distribution, Chemical Engineering Research and Design, 139: 283-295.

BIOMASS FUELS↗
DOE OSTI2017

EpiK: A Knowledge Base for Epidemiological Modeling and Analytics of Infectious Diseases

Computational epidemiology seeks to develop computational methods to study the distribution and determinants of health-related states or events (including disease), and the application of this study to the control of diseases and other health problems. Recent advances in computing and data sciences have led to the development of innovative modeling environments to support this important goal. The datasets used to drive the dynamic models as well as the data produced by these models presents unique challenges owing to their size, heterogeneity and diversity. These datasets form the basis of effective and easy to use decision support and analytical environments. As a result, it is important to develop scalable data management systems to store, manage and integrate these datasets. In this paper, we develop EpiK —a knowledge base that facilitates the development of decision support and analytical environments to support epidemic science. An important goal is to develop a framework that links the input as well as output datasets to facilitate effective spatio-temporal and social reasoning that is critical in planning and intervention analysis before and during an epidemic. The data management framework links modeling workflow data and its metadata using a controlled vocabulary. The metadata captures information about storage, the mapping between the linked model and the physical layout, and relationships to support services. EpiK is designed to support agent-based modeling and analytics frameworks—aggregate models can be seen as special cases and are thus supported. We use semantic web technologies to create a representation of the datasets that encapsulates both the location and the schema heterogeneity. The choice of RDF as a representation language is motivated by the diversity and growth of the datasets that need to be integrated. A query bank is developed—the queries capture a broad range of questions that can be posed and answered during a typical case study pertaining to disease outbreaks. The queries are constructed using SPARQL Protocol and RDF Query Language (SPARQL) over the EpiK. EpiK can hide schema and location heterogeneity while efficiently supporting queries that span the computational epidemiology modeling pipeline: from model construction to simulation output. As a result, we show that the performance of benchmark queries varies significantly with respect to the choice of hardware underlying the database and resource description framework (RDF) engine.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2014

TU-F-18C-05: Evaluation of a Method to Calculate Patient-Oriented MGD Coefficients Using Estimates of Glandular Tissue Distribution

Purpose: Evaluate mammographic mean glandular dose (MGD) coefficients for particular known tissue distributions using a novel formalism that incorporates the effect of the heterogeneous glandular tissue distribution, by comparing them with MGD coefficients derived from the corresponding anthropomorphic computer breast phantom. Methods: MGD coefficients were obtained using MCNP5 simulations with the currently used homogeneous assumption and the heterogeneously-layered breast (HLB) geometry and compared against those from the computer phantom (ground truth). The tissue distribution for the HLB geometry was estimated using glandularity map image pairs corrected for the presence of non-glandular fibrous tissue. Heterogeneity of tissue distribution was quantified using the glandular tissue distribution index, Idist. The phantom had 5 cm compressed breast thickness (MLO and CC views) and 29% whole breast glandular percentage. Results: Differences as high as 116% were found between the MGD coefficients with the homogeneous breast core assumption and those from the corresponding ground truth. Higher differences were found for cases with more heterogeneous distribution of glandular tissue. The Idist for all cases was in the [−0.8{sup −}+0.3] range. The use of the methods presented in this work results in better agreement with ground truth with an improvement as high as 105 pp. The decrease in difference across all phantom cases was in the [9{sup −}105] pp range, dependent on the distribution of glandular tissue and was larger for the cases with the highest Idist values. Conclusion: Our results suggest that the use of corrected glandularity image pairs, as well as the HLB geometry, improves the estimates of MGD conversion coefficients by accounting for the distribution of glandular tissue within the breast. The accuracy of this approach with respect to ground truth is highly dependent on the particular glandular tissue distribution studied. Predrag Bakic discloses current funding from NIH, NSF, and DoD, former funding from Real Time Tomography, LLC and a current research collaboration with Barco and Hologic.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2023

Computational and Experimental Characterization of the Ligand Environment of a Ni-Oxo Catalyst Supported in the Metal–Organic Framework NU-1000

Heterogeneous catalysts exhibit significant changes in composition due to the influence of operating conditions, and these compositional changes can have dramatic effects on catalytic performance. For traditional bulk metal heterogeneous catalysts, relationships between composition and catalytic operating conditions are well documented. However, the influence of operating conditions on the compositions of single-site heterogeneous catalysts remains largely unresolved. To address this, we report a combined computational and experimental characterization of a Ni oxo catalyst under catalytic hydrogenation conditions. Specifically, pair distribution function (PDF) analysis is combined with ab initio thermodynamic modeling to investigate ligand environments present on a Ni oxo cluster supported in the metal–organic framework NU-1000. Comparisons of the experimentally observed and simulated Ni–O coordination numbers and Ni–O, Ni···Ni, and Ni···Zr distances provide insight into the Ni ligand environment under H 2 (g). These comparisons suggest significant OH and H 2 O content and, further, that different Ni ions within the cluster and/or NU-1000 structure may comprise subtly different numbers of these ligands. Further, the observation of significant H 2 O content under H 2 (g) suggests that the NU-1000 support supplies H 2 O to the cluster. Examples of ligand environments that could lead to the observed PDFs are provided. Furthermore, the combination of simulations and experiments provides new insights into the ligand environment for Ni-NU-1000 catalysts that will be useful for understanding the ligand environments of other single-site Ni catalysts as well.

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

Computing the Properties of Matter with Leadership Computing Resources (Closeout Report for DE-SC0018121)

In order to add more capabilities to Halide, we have designed a new framework called Tiramisu and integrated this framework into Halide. Since Tiramisu enables Halide to target heterogeneous architectures, our development efforts have been refocused on Tiramisu. Most high-performance computer systems today are complex and increasingly heterogeneous; they may have CPUs, GPUs and FPGAs. Achieving best performance requires taking full advantage of all these different architectures. To address this issue, we have designed Tiramisu, an optimization framework that enables Halide (and other DSLs) to target heterogeneous architectures. Tiramisu is an optimization framework that takes as input a high level, architecture-independent representation of code and a set of scheduling and data mapping commands that guide code transformation. The input can either be generated by a domain-specific language (DSL) compiler such as Halide or directly written by a programmer. Tiramisu then applies the user-specified code and data-layout transformations and generates an architecture-specific, low-level intermediate representation (IR) that takes advantage of modern architectural features such as multicore parallelism, non-uniform memory (NUMA) hierarchies, clusters, and accelerators like GPUs and FPGAs. We integrated Tiramisu within Halide and implemented a representative set of benchmarks to evaluate this integration. Tiramisu is now open source and is available for public use (http://tiramisu-compiler.org/). A paper about Tiramisu was published, it shows that Tiramisu extends Halide with many new capabilities and that Tiramisu can generate efficient code for multicores, GPUs, FPGAs and distributed heterogeneous systems. The performance of code generated by the Tiramisu backends matches or exceeds hand optimized reference implementations. For example, the multicore backend matches the highly optimized Intel MKL library on many kernels and shows speedups reaching 4x over the original Halide. In addition to making Tiramisu more robust, we have used Tiramisu to implement a set of representative tensor operation for constructing baryon building blocks required for multi baryon contractions in LQCD. In order to implement this code, we needed to generalize Tiramisu in two ways: first we needed to support indirect array accesses, and second, we needed to add support for complex numbers to Tiramisu. The code generated by Tiramisu is 6x faster than the reference code. Our efforts towards an MPI based multi-node version of tiramisu have matured and the resulting code scales well on multiple nodes (tests up to 512 KNL nodes have been undertaken).

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2012

Comparison of Different Upscaling Methods for Predicting Thermal Conductivity of Complex Heterogeneous Materials System: Application on Nuclear Waste Forms

To develop a strategy in thermal conductivity prediction of a complex heterogeneous materials system, loaded nuclear waste forms, the computational efficiency and accuracy of different upscaling methods have been evaluated. The effective thermal conductivity, obtained from microstructure information and local thermal conductivity of different components, is critical in predicting the life and performance of waste form during storage. Several methods, including the Taylor model, Sachs model, self-consistent model, and statistical upscaling method, were developed and implemented. Microstructure based finite element method (FEM) prediction results were used to as benchmark to determine the accuracy of the different upscaling methods. Micrographs from waste forms with varying waste loadings were used in the prediction of thermal conductivity in FEM and homogenization methods. Prediction results demonstrated that in term of efficiency, boundary models (e.g., Taylor model and Sachs model) are stronger than the self-consistent model, statistical upscaling method, and finite element method. However, when balancing computational efficiency and accuracy, statistical upscaling is a useful method in predicting effective thermal conductivity for nuclear waste forms.

36 MATERIALS SCIENCE↗
DOE OSTI2018

Extracting Knowledge from Data through Catalysis Informatics

Catalysis informatics is a distinct sub-field that lies at the intersection of cheminformatics and materials informatics, but with unique challenges arising from the dynamic, surface-sensitive, multi-scale nature of heterogeneous catalysis. The ideas behind catalysis informatics can be traced back decades, but the field is only recently emerging due to advances in data infrastructure, statistics, machine learning, and computational methods. In this work we review the field from early works on expert systems and knowledge engines to more recent approaches utilizing machine-learning and uncertainty quantification. The data-information-knowledge hierarchy is introduced and used to classify various developments. The chemical master equation and micro-kinetic models are proposed as a quantitative representation of catalysis knowledge, which can be used to generate explanative and predictive hypotheses for the understanding and discovery of catalytic materials. We discuss future prospects for the field, including improved quantitative coupling of experiment/theory, advanced micro-kinetic models, and the development of open-source software tools. As a result, integration of existing chemical and physical models with emerging statistical and computational tools presents a promising route toward the automated design, discovery, and optimization of heterogeneous catalytic processes.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2018

Prediction of Adsorption Energies for Chemical Species on Metal Catalyst Surfaces Using Machine Learning

Computational catalyst screening has the potential to significantly accelerate heterogeneous catalyst discovery. Typically, this involves developing microkinetic reactor models that are based on parameters obtained from density functional theory and transition-state theory. To reduce the large computational cost involved in computing various adsorption and transition-state energies of all possible surface states on a large number of catalyst models, linear scaling relations for surface intermediates and transition states have been developed that only depend on a few, typically one or two descriptors, such as the carbon atom adsorption energy. As a result, only the descriptor values have to be computed for various active site models to generate volcano curves in activity or selectivity. Unfortunately, for more complex chemistries the predictability of linear scaling relations is unknown. Also, the selection of descriptors is essentially a trial and error process. Here, using a database of adsorption energies of the surface species involved in the decarboxylation and decarbonylation of propionic acid over eight monometalic transition-metal catalyst surfaces (Ni, Pt, Pd, Ru, Rh, Re, Cu, Ag), we tested if nonlinear machine learning (ML) models can outperform the linear scaling relations in prediction accuracy when predicting the adsorption energy for various species on a metal surface based on data from the rest of the metal surfaces. We found linear scaling relations to hold well for predictions across metals with a mean-absolute error of 0.12 eV, and ML methods being unable to outperform linear scaling relations when the training dataset contains a complete set of energies for all of the species on various metal surfaces. Only when the training dataset is incomplete, namely, contains a random subset of species’ energies for each metal, a currently unlikely scenario for catalyst screening, do kernel-based ML models significantly outperform linear scaling relations. We also found that simple coordinate-free species descriptors, such as bond counts, achieve as good results as sophisticated coordinate-based descriptors. Finally, we propose an approach for automatic discovery of appropriate metal descriptors using principal component analysis.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Celeritas R&D Report: Accelerating Geant4

Celeritas is a new Monte Carlo (MC) detector simulation code designed for computationally intensive applications on high-performance heterogeneous architectures. In the past two years Celeritas has advanced from prototyping a Graphics Processing Unit (GPU)-based single physics model in infinite medium to implementing a full set of electromagnetic (EM) physics processes in complex geometries. The current release of Celeritas, version 0.4, has incorporated full device-based navigation, an event loop in the presence of magnetic fields, and detector hit scoring. New functionality incorporates a scheduler to offload electromagnetic physics to the GPU within a Geant4-driven simulation, enabling straightforward integration of Celeritas into the high energy physics (HEP) experimental frameworks CMSSW and ATLAS FullSimLight. On the Perlmutter supercomputer, Celeritas performs EM physics between 3× and 18× faster using the machine’s Nvidia GPUs compared to using only CPUs, corresponding to an electrical power efficiency up to a factor of 5. When running a multithreaded Geant4 ATLAS test beam application with full hadronic physics, using Celeritas to accelerate the EM physics results in an overall simulation speedup of 1.7–2.2× on GPU and 1.2× on CPU. In a CMS test application using tt¯ events and the prototype Run 4 configuration, compared to Geant4 CPU, Celeritas with a Nvidia A100 improves overall throughput up to a factor of 2.7× but cannot be efficiently shared with more than 8 cores.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2024

Iterative methods in GPU-resident linear solvers for nonlinear constrained optimization

Linear solvers are major computational bottlenecks in a wide range of decision support and optimization computations. The challenges become even more pronounced on heterogeneous hardware, where traditional sparse numerical linear algebra methods are often inefficient. For example, methods for solving ill-conditioned linear systems have relied on conditional branching, which degrades performance on hardware accelerators such as graphical processing units (GPUs). To improve the efficiency of solving ill-conditioned systems, our computational strategy separates computations that are efficient on GPUs from those that need to run on traditional central processing units (CPUs). Our strategy maximizes the reuse of expensive CPU computations. Iterative methods, which thus far have not been broadly used for ill-conditioned linear systems, play an important role in our approach. In particular, we extend ideas from Arioli et al., (2007) to implement iterative refinement using inexact LU factors and flexible generalized minimal residual (FGMRES), with the aim of efficient performance on GPUs. In conclusion, we focus on solutions that are effective within broader application contexts, and discuss how early performance tests could be improved to be more predictive of the performance in a realistic environment.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

Scalable Computation of 2D-Minkowski Sum of Arbitrary Non-Convex Domains: Modeling Flexibility in Energy Resources

The flexibility of active ($p$) and reactive power ($q$) consumption in distributed energy resources (DERs) can be represented as a (potentially non-convex) set of points in the $p$-$q$ plane. Modeling of the aggregated flexibility in a heterogeneous ensemble of DERs as a Minkowski sum (M-sum) is computationally intractable even for moderately sized populations. In this article, we propose a scalable method of computing the M-sum of the flexibility domains of a heterogeneous ensemble of DERs, which are allowed to be non-convex, non-compact. In particular, the proposed algorithm computes a guaranteed superset of the true M-sum, with desired accuracy. The worst-case complexity of the algorithm is computed. Special cases are considered, and it is shown that under certain scenarios, it is possible to achieve a complexity that is linear with the size of the ensemble. Numerical examples are provided by computing the aggregated flexibility of different mix of DERs under varying scenarios.

Kundu, Soumya↗
DOE OSTI2019

Highly-scalable, physics-informed GANs for learning solutions of stochastic PDEs

Uncertainty quantification for forward and inverse problems is a central challenge across physical and biomedical disciplines. We address this challenge for the problem of modeling subsurface flow at the Hanford Site by combining stochastic computational models with observational data using physics-informed GAN models. The geographic extent, spatial heterogeneity, and multiple correlation length scales of the Hanford Site require training a computationally intensive GAN model to thousands of dimensions. We develop a highly optimized implementation that scales to {\herogpucount} NVIDIA Volta GPUs. We develop a hierarchical scheme based on a multi-player game-theoretic approach for exploiting domain parallelism, map discriminators and generators to multiple GPUs, and employ efficient communication schemes to ensure training stability and convergence. Our implementation scales to {\heronodecount} nodes on the Summit supercomputer with a \weakscalingeff\% scaling efficiency, achieving peak and sustained half-precision rates of {\heropeakperf} PF/s and {\herosustainedperf} PF/s.

physics informed ML, machine learning, generative ↗
DOE OSTI2020

Contributions to Lignomics: Stochastic Generation of Oligomeric Lignin Structures for Interpretation of MALDI–FT‐ICR‐MS Results

The lack of standards to identify oligomeric molecules is a challenge for the analysis of complex organic mixtures. High-resolution mass spectrometry—specifically, Fourier-transform ion cyclotron resonance mass spectrometry (FT-ICR MS)—offers new opportunities for analysis of oligomers with the assignment of formulae (C x H y O z ) to detected peaks. However, matching a specific structure to a given formula remains a challenge due to the inability of FT-ICR MS to distinguish between isomers. Additional separation techniques and other analyses (e.g., NMR spectroscopy) coupled with comparison of results to those from pure compounds is one route for assignment of MS peaks. Unfortunately, this strategy may be impractical for complete analysis of complex, heterogeneous samples. In this study we use computational stochastic generation of lignin oligomers to generate a molecular library for supporting the assignment of potential candidate structures to compounds detected during FT-ICR MS analysis. Furthermore, this approach may also be feasible for other macromolecules beyond lignin.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2018

Numerical study of multiscale compaction-initiated detonation

A multiscale model of heterogeneous condensed-phase explosives is examined computationally to determine the course of transient events following the application of a piston-driven stimulus. The model is a modified version of that introduced by Gonthier (Combust Sci Technol 175(9):1679–1709, 2003. https://doi.org/10.1080/00102200302373) in which the explosive is treated as a porous, compacting medium at the macro-scale and a collection of closely packed spherical grains capable of undergoing reaction and diffusive heat transfer at the meso-scale. A separate continuum description is ascribed to each scale, and the two scales are coupled together in an energetically consistent manner. Following piston-induced compaction, localized energy deposition at the sites of intergranular contact creates hot spots where reaction begins preferentially. Reaction progress at the macro-scale is determined by the spatial average of that at the grain scale. A parametric study shows that combustion at the macro-scale produces an unsteady detonation with a cyclical character, in which the lead shock loses strength and is overtaken by a stronger secondary shock generated in the partially reacted material behind it. Finally, the secondary shock in turn becomes the new lead shock and the process repeats itself.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2022

Physics-informed machine learning with differentiable programming for heterogeneous underground reservoir pressure management

Abstract Avoiding over-pressurization in subsurface reservoirs is critical for applications like CO $$_2$$ 2 sequestration and wastewater injection. Managing the pressures by controlling injection/extraction are challenging because of complex heterogeneity in the subsurface. The heterogeneity typically requires high-fidelity physics-based models to make predictions on CO $$_2$$ 2 fate. Furthermore, characterizing the heterogeneity accurately is fraught with parametric uncertainty. Accounting for both, heterogeneity and uncertainty, makes this a computationally-intensive problem challenging for current reservoir simulators. To tackle this, we use differentiable programming with a full-physics model and machine learning to determine the fluid extraction rates that prevent over-pressurization at critical reservoir locations. We use DPFEHM framework, which has trustworthy physics based on the standard two-point flux finite volume discretization and is also automatically differentiable like machine learning models. Our physics-informed machine learning framework uses convolutional neural networks to learn an appropriate extraction rate based on the permeability field. We also perform a hyperparameter search to improve the model’s accuracy. Training and testing scenarios are executed to evaluate the feasibility of using physics-informed machine learning to manage reservoir pressures. We constructed and tested a sufficiently accurate simulator that is 400 000 times faster than the underlying physics-based simulator, allowing for near real-time analysis and robust uncertainty quantification.

54 ENVIRONMENTAL SCIENCES↗