Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “deep well”

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 253 records · Page 14
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Discrete Graph Structure Learning for Forecasting Multiple Time Series

Time series forecasting is an extensively studied subject in statistics, economics, and computer science. Exploration of the correlation and causation among the variables in a multivariate time series shows promise in enhancing the performance of a time series model. When using deep neural networks as forecasting models, we hypothesize that exploiting the pairwise information among multiple (multivariate) time series also improves their forecast. If an explicit graph structure is known, graph neural networks (GNNs) have been demonstrated as powerful tools to exploit the structure. In this work, we propose learning the structure simultaneously with the GNN if the graph is unknown. We cast the problem as learning a probabilistic graph model through optimizing the mean performance over the graph distribution. The distribution is parameterized by a neural network so that discrete graphs can be sampled differentiably through reparameterization. Empirical evaluations show that our method is simpler, more efficient, and better performing than a recently proposed bilevel learning approach for graph structure learning, as well as a broad array of forecasting models, either deep or non-deep learning based, and graph or non-graph based.

Shang, Chao↗
DOE OSTI2023

Daytime convective development over land: the role of surface forcing

Water availability at the Earth surface determines the portioning of the surface heat flux into its sensible and latent components, that is, the surface heat flux Bowen ratio. The two components affect differently the surface buoyancy flux and thus the development and growth of the convective boundary layer. As a result, the Bowen ratio has a strong impact on the daytime dry and moist convection development over land. Here, we use two canonical modeling test cases, one for the shallow convection and one for the shallow-to-deep convection transition, to document the impact of the surface Bowen ratio on daytime convection development. A simple approach is used where results from simulations featuring the original setup are contrasted with simulations where surface sensible heat flux takes on values of the latent heat flux and vice versa. Such a change illustrates the key impact of the surface water availability without changing the total surface heat flux. Because of the larger surface buoyancy flux, simulations with the reversed surface heat fluxes feature faster deepening of the convective boundary layer and wider clouds once moist convection develops. Mean cloud base width of cumulus clouds increases with the boundary layer depth. For shallow-to-deep convection transition, comparison of simulations with the observed large-scale flow that feature substantial lower-tropospheric shear with simulations with no mean horizontal flow suggests that the organization of convective circulations within the boundary layer strongly affects initial size and organization of shallow convection, as well as the subsequent width increase once convection deepens and eventually transitions to deep precipitating convection. A simple explanation is provided of why deeper well-mixed convective subcloud layer results in wider clouds. The key is the larger width of boundary layer coherent updraft structures when the mixed layer is deeper.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Groundwater characterization and monitoring at a complex industrial waste site using electrical resistivity imaging

A contaminated industrial waste site in Washington State (USA) containing buried, metallic-waste storage tanks, pipes, and wells, was evaluated to determine the feasibility of monitoring groundwater remediation activities associated with an underlying perched aquifer system using electrical resistivity tomography. The perched aquifer, located ~65 m below ground surface and ~10 m above the regional water table, contains high concentrations of nitrate, uranium, and other contaminants of concern from past tank leaks and intentional releases of wastes to surface disposal sites. The extent of the perched water aquifer is not well known, and the effectiveness of groundwater extraction for contaminant removal is uncertain, so supplemental characterization and monitoring technologies are being evaluated. Numerical simulations of subsurface flow and contaminant transport were performed with a highly resolved model of the hydrogeologic system and waste site infrastructure, and these simulations were used as the physical basis for electrical resistivity tomography modeling. The modeling explicitly accounted for metallic infrastructure at the site. The effectiveness of using surface electrodes versus surface and horizontal subsurface electrodes, for imaging groundwater extraction from the perched water aquifer, was investigated. Although directional drilling is a mature technology, its use for electrode emplacement in the deep subsurface under a complex industrial waste site via horizontal wells has not yet been demonstrated. Results from this study indicate that using horizontal subsurface electrode arrays could significantly improve the ability of electrical resistivity tomography to image deep subsurface features and monitor remediation activities under complex industrial waste sites.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

NRAP-Open-IAM: Generic Aquifer Component Development and Testing

The Generic Aquifer Model calculates the concentrations of dissolved salt and dissolved CO 2 surrounding a leaking legacy well. The Generic Aquifer model can also estimate the size of an “impact plume” where concentration changes exceed user-specified thresholds. The model is a component of NRAP-Open-IAM, an open-source Integrated Assessment Model (IAM) developed by the National Risk Assessment Partnership (NRAP) to perform risk assessment for geologic CO 2 storage. The input parameters were selected to cover a wide range of groundwater aquifers and leakage rates. The generic aquifer model was developed using a generative adversarial deep learning network, trained using a large synthetic dataset of STOMP multiphase flow simulations. The deep learning model predictions of dissolved salt and dissolved CO 2 in the aquifer compare well to the original STOMP simulation results. The extent of aquifer impacted by leaking CO 2 or brine is calculated using a user-defined mass fraction threshold. The aquifer impact volumes calculated based on STOMP simulation results compare well to those calculated based on the deep learning model. In a provided python script, gridded observation results from the generic aquifer component of NRAP-Open-IAM are converted to HDF5 format files for monitoring design with the DREAM code.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Deep Eutectic Solvents as Candidates for Lithium Isotope Enrichment

Nuclear fusion is a phenomenon that is well known within the nuclear physics community as a viable option for alternative energy as many natural gases and fossil fuels are phased out of commercial use. Deuterium and tritium fusion reactions are currently the leading candidates for nuclear fusion, with a major limiting factor being a means to produce tritium on an industrial scale. Lithium-6 is a well-known isotope that can produce tritium and helium following a fission reaction with a neutron. Unfortunately, the lithium-6 enrichment methods are limited to the COLEX process, which leaves behind an alarming amount of mercury waste as a potential environmental contaminant. Deep eutectic solvents are believed to be a potential alternative to lithium isotope separations due to the ease of generation, in addition to the minimum environmental waste generated when these solvents are employed. Previous studies have suggested that deep eutectic solvents are capable of separating lithium isotopes by utilizing a 2-thenoyltrifluoroacetone and trioctylphosphine oxide system that can biphasically react with a buffered solution containing lithium chloride. This system displays a separation factor of 1.068, which when compared to the 1.054 separation within the COLEX process, makes it a potential candidate for lithium-6/7 separation. Within this study, we investigate this system in comparison to two newly synthesized deep eutectic solvents and find that within these acetylacetone-based systems, little isotopic separation is observed. We investigate these systems both experimentally and computationally, showing the different lithium cation affinities, in addition to proposing how the electron-donating or -withdrawing nature can influence these systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2026

Tuning the Solvation and Solubility Properties of Molecularly Heterogeneous Nonionic Deep Eutectic Solvents via Interface Organization

Common separation techniques, such as liquid− liquid extraction, are usually used for extractions and purifications due to their industrial scalability and affordability. However, these well-established practices are hindered by low selectivity and challenges in recovering solutes and solvents. Deep eutectic solvents (DES), a fairly new type of solvent, have the potential to overcome these issues. DESs are binary mixtures whose physical properties can be tuned by selecting the appropriate precursors to facilitate and/or enhance processes such as extraction. A promising DES for selective separations is formed when lauric acid (LA) is mixed with N-methylacetamide (NMA). This LA-NMA DES has a heterogeneous microscopic structure that can solvate compounds with completely different polarities. This study explores how forming an organized structure on the mesoscale affects the solubility of nonpolar solutes in nonionic DESs. To this end, the molecular and mesoscale structures and their effect on the solubility and solvation properties are evaluated for the LA-NMA DES and two new DESs with slight chemical variations in their precursors. It is observed that the organization of the nonpolar DES domains and, consequently, of their interfaces directly relates to the solubility of nonpolar compounds. Specifically, correctly selecting the DES precursors that form organized nonpolar domains leads to an organized interface in which the nonpolar solutes are solvated, thereby increasing the solubility. Additionally, enhanced dissolution power was observed in a completely different DES with mesoscale order in its molecular structure and composed of menthol and lauric acid. The latter result further validates the proposed tunability of the DES dissolution power through organized interfaces, extending it beyond a specific DES family and opening the possibility of new extraction-tailored designer solvents.

Extraction↗
DOE OSTI2021

Deep Learning Estimation of Daily Ground–Level NO 2 Concentrations from Remote Sensing Data

The limited number of nitrogen dioxide (NO 2 ) surface measurements calls for the development of highly accurate approaches to estimating surface NO 2 concentrations. In this study, we leverage a new satellite instrument, the TROPOspheric Monitoring Instrument (TROPOMI), along with other predictor variables, to estimate daily surface NO 2 concentrations over Texas in 2019. We use the deep convolutional neural network (Deep-CNN), an advanced deep learning algorithm, to obtain estimates and achieve a correlation coefficient (R) of 0.91, an index of agreement (IOA) of 0.95, and a mean absolute bias (MAB) of 1.75 ppb in surface NO 2 estimation. Additionally, we leverage a novel approach, SHapley Additive exPlanations (SHAP), to describe how Deep-CNN understands each predictor variable. The SHAP results show that the Deep-CNN model has an advanced understanding of the dataset, revealing that TROPOMI closely captures levels of NO 2 . In addition, we show the superiority of our Deep-CNN model at estimating surface NO 2 over other well-known machine learning and regression models in the field, including the support vector machines (SVM), random forest (RF), and multiple linear regression (MLR). Although SVM and RF show strong capabilities at estimating surface NO 2 concentrations, their accuracy is inferior to that of the Deep-CNN model, ranking second and third in model accuracy in this study. The MLR, however, shows a poor ability at NO 2 estimation and ranks last among all models. Furthermore, testing the impact of sample size on model performance, we also show that, compared to other models, Deep-CNN needs more samples to trigger its strength at surface NO 2 estimation.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Equation of State and Spin Crossover of (Al, Fe)–Phase H

The transport of hydrogen into Earth's deep interior may have an impact on lower mantle dynamics as well as on the seismic signature of subducted material. Due to the stability of the hydrous phases δ-AlOOH (delta phase), MgSiO 2 (OH) 2 (phase H), and ε-FeOOH at high temperatures and pressures, their solid solutions may transport significant amounts of hydrogen as deep as the core-mantle boundary. Here we have constrained the equation of state, including the effects of a spin crossover in the Fe 3+ atoms, of (Al, Fe)-phase H: Al 0.84 Fe 3+ 0.07 Mg 0.02 Si 0.06 OOH, using powder X-ray diffraction measurements to 125 GPa, supported by synchrotron Mössbauer spectroscopy measurements on (Al, Fe)-phase H and δ-(Al, Fe)OOH. The changes in spin state of Fe 3+ in (Al, Fe)-phase H results in a significant decrease in bulk sound velocity and occurs over a different pressure range (48–62 GPa) compared with δ-(Al, Fe)OOH (32–40 GPa). Changes in axial compressibilities indicate a decrease in the compressibility of hydrogen bonds in (Al, Fe)-phase H near 30 GPa, which may be associated with hydrogen bond symmetrization. The formation of (Al, Fe)-phase H in subducted oceanic crust may contribute to scattering of seismic waves in the mid-lower mantle (~1,100–1,550 km). Accumulation of 1–4 wt.% (Al, Fe)-phase H could reproduce some of the seismic signatures of large, low seismic-velocity provinces. Our results suggest that changes in the electronic structure of phases in the (δ-AlOOH)-(MgSiO 2 (OH) 2 )-(ε-FeOOH) solid solution are sensitive to composition and that the presence of these phases in subducted oceanic crust could be seismically detectable throughout the lower mantle.

58 GEOSCIENCES↗
DOE OSTI2022

Uncertainty Quantification Framework for Predicting Material Response with Large Number of Parameters: Application to Creep Prediction in Ferritic-Martensitic Steels Using Combined Crystal Plasticity and Grain Boundary Models

This paper presents an uncertainty quantification (UQ) framework for the physics-based model prediction of material response with a large number of parameters. The application problem presented in this work is that of predicting creep in Grade 91 steel at 600°C. The material response is defined with a physically based microstructural model with constitutive equations emulating several observed phenomena in Grade 91 and embodied into an explicit geometry mesoscale finite element model for prior austenite grains and grain boundaries. Creep within the grains and in grain boundaries are represented by crystal plasticity for dislocation motion and a physics-based model for cavity growth and nucleation, respectively. The creep behavior of this material is influenced by several parameters, some of which have a wide range of variation based on experimental data. UQ combined with microstructural modeling can discover the core microstructural causes of experimental variability, leading to improved materials with lower variability in critical long-term material properties. In this study, we investigate the model's uncertainty to identify material properties that may be modified during production to increase creep life and analyze different components of the crystal plasticity model for improvements. For this purpose, a quantity of interest is defined as time to minimum creep rate, which correlates well to the creep failure of the material. A deep neural network model was trained and validated to be used as a surrogate for the finite element model. Then, a variance-based sensitivity analysis is performed on the surrogate model to find the Sobol indices of the input parameters in respect to the output quantity of interest. The Sobol indices are used to reduce the dimensionality of the model. Generalized polynomial chaos expansion is used on the reduced basis models to propagate the uncertainty from the input parameters to the quantity of interest using the deep neural network surrogate model. These results are benchmarked against uncertainty propagation using Monte Carlo simulations. In conclusion, the UQ performed through the reduced basis model captures almost all the uncertainty in the model with significantly fewer simulations, making it possible to perform the UQ directly via simulations with the finite element model rather than surrogate machine-learned models.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Phylogenomics reveals convergent evolution of red-violet coloration in land plants and the origins of the anthocyanin biosynthetic pathway

The flavonoids, one of the largest classes of plant secondary metabolites, are found in lineages that span the land plant phylogeny and play important roles in stress responses and as pigments. Perhaps the most well-studied flavonoids are the anthocyanins that have human health benefits and help plants attract pollinators, regulate hormone production, and confer resistance to abiotic and biotic stresses. The canonical biochemical pathway responsible for the production of these pigments is well-characterized for flowering plants yet its conservation across deep divergences in land plants remains debated and poorly understood. Many early land plants such as mosses, liverworts, and ferns produce flavonoid pigments, but their biosynthetic origins and homologies to the anthocyanin pathway remain uncertain. We conducted phylogenetic analyses using full genome sequences representing nearly all major green plant lineages to reconstruct the evolutionary history of the anthocyanin biosynthetic pathway then test the hypothesis that genes in this pathway are present in early land plants. We found that the entire pathway was not intact until the most recent common ancestor of seed plants and that orthologs of many downstream enzymes are absent from seedless plants including mosses, liverworts, and ferns. Our results also highlight the utility of phylogenetic inference, as compared to pairwise sequence similarity, in orthology assessment within large gene families that have complex duplication-loss histories. Here, we suggest that the production of red-violet flavonoid pigments widespread in seedless plants, including the 3-deoxyanthocyanins, requires the activity of novel, as-yet discovered enzymes, and represents convergent evolution of red-violet coloration across land plants.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

Facile One-Pot Nanoproteomics for Label-Free Proteome Profiling of 50–1000 Mammalian Cells

Recent advances in sample preparation enable label-free MS-based proteome profiling of small numbers of mammalian cells. However, specific devices are often required to downscale sample processing volume from the standard 50-200 µL to sub-µL for effective nanoproteomics, which greatly impedes the implementation of current nanoproteomics methods by broad proteomics research community. Here we report a facile one-pot nanoproteomics method termed SOPs-MS (Surfactant-assisted One-Pot sample processing at the standard volume coupled with MS) for convenient proteome profiling of 50-1000 mammalian cells. Building upon our recent development of SOP-MS for label-free single-cell proteomics at low µL volume (Commun Bio 2021, 4, 265), we have systematically evaluated its processing volume at 10-200 µL using 100 human cells for robust reproducible nanoproteomic analysis. The processing volume of 50 µL which is in the range of volume for standard proteomics sample preparation, has been selected for easy sample handling with benchtop micropipette. Using the commonly accessible LC-MS platform, SOPs-MS allows for reliable label-free quantification of ~1200-2700 protein groups from 50-1000 MCF10A cells. When applied to small subpopulations of mouse colon crypt cells, SOPs-MS can reveal distinct protein signatures between any two subpopulation cells with identification of ~1500-2500 protein groups for each subpopulation. SOPs-MS may pave the way for routine deep proteome profiling of small numbers of cells as well as low-input samples.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2023

Explainable deep learning for insights in El Niño and river flows

The El Niño Southern Oscillation (ENSO) is a semi-periodic fluctuation in sea surface temperature (SST) over the tropical central and eastern Pacific Ocean that influences interannual variability in regional hydrology across the world through long-range dependence or teleconnections. Recent research has demonstrated the value of Deep Learning (DL) methods for improving ENSO prediction as well as Complex Networks (CN) for understanding teleconnections. However, gaps in predictive understanding of ENSO-driven river flows include the black box nature of DL, the use of simple ENSO indices to describe a complex phenomenon and translating DL-based ENSO predictions to river flow predictions. Here we show that eXplainable DL (XDL) methods, based on saliency maps, can extract interpretable predictive information contained in global SST and discover SST information regions and dependence structures relevant for river flows which, in tandem with climate network constructions, enable improved predictive understanding. Our results reveal additional information content in global SST beyond ENSO indices, develop understanding of how SSTs influence river flows, and generate improved river flow prediction, including uncertainty estimation. Observations, reanalysis data, and earth system model simulations are used to demonstrate the value of the XDL-CN based methods for future interannual and decadal scale climate projections.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Identification and reconstruction of low-energy electrons in the ProtoDUNE-SP detector

Measurements of electrons from 𝜈𝑒 interactions are crucial for the Deep Underground Neutrino Experiment (DUNE) neutrino oscillation program, as well as searches for physics beyond the standard model, supernova neutrino detection, and solar neutrino measurements. This article describes the selection and reconstruction of low-energy (Michel) electrons in the ProtoDUNE-SP detector. ProtoDUNE-SP is one of the prototypes for the DUNE far detector, built and operated at CERN as a charged particle test beam experiment. A sample of low-energy electrons produced by the decay of cosmic muons is selected with a purity of 95%. This sample is used to calibrate the low-energy electron energy scale with two techniques. An electron energy calibration based on a cosmic ray muon sample uses calibration constants derived from measured and simulated cosmic ray muon events. Another calibration technique makes use of the theoretically well-understood Michel electron energy spectrum to convert reconstructed charge to electron energy. In addition, the effects of detector response to low-energy electron energy scale and its resolution including readout electronics threshold effects are quantified. Finally, the relation between the theoretical and reconstructed low-energy electron energy spectra is derived, and the energy resolution is characterized. The low-energy electron selection presented here accounts for about 75% of the total electron deposited energy. After the addition of lost energy using a Monte Carlo simulation, the energy resolution improves from about 40% to 25% at 50 MeV. These results are used to validate the expected capabilities of the DUNE far detector to reconstruct low-energy electrons.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2020

Research Introduction [Slides]

In the topic of Reverse Time Imaging, we proposed a new IC to reduce computation cost but reserve image resolution for distributed sensor networks. For Induced Seismicity in Oklahoma, we analyzed fault stress state analysis at state scale, and we applied machine learning techniques to polarity picking and seismicity rate forecasting. The results provide better understanding of fault properties, stress field, and the relationships among fault, stress state, injections, and potential seismic hazards. Lastly, for Microseismic Monitoring, we detected and located 770 low-frequency events (5-50 Hz): (1) Shallow events are highly clustered, consistent with the pathway from injection well 13-10A to monitoring well; moment tensor inversion shows dominant tensile cracking; (2) Deep events are scattered and show migration pattern to the basement; moment tensors show that most events are shear cracks.

58 GEOSCIENCES↗
DOE OSTI2025

The NGC 3109 Satellite System: The First Systematic Resolved Search for Dwarf Galaxies Around an SMC-mass Host

We report the results of the deepest search to date for dwarf galaxies around NGC 3109, a barred spiral galaxy with a mass similar to that of the Small Magellanic Cloud (SMC), using a semiautomated search method. Using the Dark Energy Camera, we survey a region covering a projected distance of ∼70 kpc of NGC 3109 (D = 1.3 Mpc, R vir ∼ 90 kpc, M ∼ 10 8 M*) as part of the MADCASH and DELVE-DEEP programs. We introduce a newly developed semiresolved search method, used alongside a resolved search, to identify crowded dwarf galaxies around NGC 3109. Using both approaches, we successfully recover the known satellites Antlia and Antlia B. We identified a promising candidate, which was later confirmed to be a background dwarf through deep follow-up observations. Our detection limits are well defined, with the sample ∼80% complete down to M V ∼ −8.0, and include detections of dwarf galaxies as faint as M V ∼ −6.0. This is the first comprehensive study of a satellite system through resolved stars around an SMC mass host. Our results show that NGC 3109 has more bright (M V ∼ −9.0) satellites than the mean predictions from cold dark matter models, but well within the host-to-host scatter. A larger sample of LMC/SMC-mass hosts is needed to test whether or not the observations are consistent with current model expectations.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

Spent Nuclear Fuel Storage and Radioactive Waste Disposal in the United States: A Law and Policy Analysis - 20302

While Congress plays political football, spent nuclear fuel continues to sit in de facto interim nuclear waste storage sites throughout the country. This ad hoc approach - if it can even be called an approach - to nuclear waste management is in no one's best interests, because of financial, safety, environmental, and national security concerns with the status quo. Senate leadership had asserted that an up-or-down vote on the Yucca Mountain Project would take place in September or October 2019, but that did not occur. Rather, a conference bill was agreed to, and it provided that, although the nuclear program should receive some funding, funding was not made available for centralized interim storage ('CIS') or the Yucca Mountain Project. The President's FY2021 Budget Request does not contain any proposed funding for the Yucca Mountain Project ('the Project'). This paper will discuss the implications of that process, both for the Project itself and for the U.S. nuclear waste program more generally. In order to provide context for the current situation, this paper will begin by chronicling the history of U.S. nuclear waste management, from the passage of the Nuclear Waste Policy Act of 1982 and to the present day. It will discuss the roles of the U.S. Department of Energy, the Nuclear Regulatory Commission, federal and state political delegations, and the nuclear energy industry. The paper will then identify key themes that can be expected to affect the future of nuclear waste management in the United States. For example, it will consider whether development of additional CIS facilities is a viable short- to medium-term solution. CIS facilities may relieve immediate pressure at decommissioned or soon-to-be-decommissioned nuclear power plants; but with that pressure partially alleviated, support for building a long-term repository may be reduced. It will also discuss opposition by the State of New Mexico and others to the proposed New Mexico CIS project, based not only on siting concerns but also whether it might become a de facto repository, and whether the State of Texas and others may also be opposed, for similar reasons, to the proposed CIS facility in Texas. Additionally, the paper will examine consent-based siting and its implications for Nevada and other states that are potential hosts for spent nuclear fuel storage and disposal facilities. Further, if the Project is not a politically viable solution, the paper will examine whether it is now time for serious consideration of an alternate location for a permanent repository, as well as consideration of other options entirely, such as deep geologic disposal. The paper will conclude by looking ahead to the 2020 elections. It will describe the positions of candidates (particularly Presidential candidates) on spent nuclear fuel storage and disposal options, without advocating for any particular candidate or political party. Perhaps in 2020, we will at least see a start to a solution to the seemingly intractable problem of nuclear waste management in the United States. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2020

Using Reinforcement Learning to Optimize Quantum Circuits in thePresence of Noise

Many quantum computing frameworks currently use noise aware algorithms for implementing quantum circuits which do not scale efficiently as the size of the hardware architecture increases. As we move towards devices which utilize more qubits, it becomes increasing more important to map quantum circuits in a way that uses resources efficiently as well as maximizes the reliability of the results of that circuit. However, as the hardware increases to the point where Quantum supremacy is attainable, it will infeasible for a brute-force algorithm to find the most optimal circuit layout for circuits of medium to large depth sizes. To this end, we will to rely on reinforcement learning (RL) as a method of building quantum circuits based on observations of the noise characteristics in its environment. In this work, we create a working reinforcement learning environment in which an agent is able to make action which will build the class of circuits which creates the GHZ state. In addition to this, we also get preliminary results of the performance of a Deep Q Neural Network, which initially does not perform as well as we believe it can. In the future, we want to improve the performance of the agent and potentially generalize this environment to more classes of circuits.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗