Engineering topics
Publications and source records attributed to Romanov, Vyacheslav.
Conference presentation on hardware acceleration of AI model training, validation, and testing using wafer scale engines, to improve efficiency and effectiveness of carbon management.
This is an invited guest speaker's presentation. The present covers three use cases by applying machine learning techniques. The use cases include fracture analysis for CCUS: IBDP study, multiple level of fracture network analysis and tool: HFTS1 study, Frac-Hit with Middleland Basin datasets from collaborations with Company A.
Saline formations deep underground are among the most promising targets for large-scale geologic carbon storage. The caprock seal integrity evaluation is an important component of commercial-scale CO 2 sequestration projects. Measurements of the porosity and permeability of mudstone samples from the NETL-supported Cranfield Project were performed using a helium porosimeter and a core flow apparatus, before and after exposure in a CO 2 -saturated brine environment. The permeability of the core samples rapidly decreased with the increase in confining pressure and did not fully recover after decompression. On the other hand, exposure to CO 2 led to an increase in the permeability by at least an order of magnitude. The porosity changes after the exposure were not substantial. The post-exposure increase in permeability was subsequently offset by its rapid decrease during the higher-pressure confinement. Extrapolation of the observed permeability trends to in situ reservoir conditions suggests that Tuscaloosa mudstone can effectively serve as a natural seal.
Invited guest speaker presentation at the 5th Annual Appalachian Basin Geophysical Symposium, June 4th, 2024, Pittsburgh, PA
The primary objective of the study was development of a machine learning (ML)-based workflow for fracture hit (“frac hit”) detection and monitoring using shale oil-field data such as drilling surveys, production history (oil and produced water), pressure, and fracking start time and duration records. The ML method takes advantage of long short-term memory (LSTM) and multilayer perceptron (MLP) neural networks to identify the frac hits due to hydraulic communication between the fracking child well(s) and the producing parent well(s) within the same pad (intra-pad interaction) and/or on different pads (inter-pad interaction). It utilizes time series of pressure and production data from within a pad and from adjacent pads. The workflow can capture time variable features of frac hits when the model architecture is deep and wide enough, with enough trainable parameters for deep learning and feature extraction, as demonstrated in this paper by using training and testing subsets of the field data from selected neighboring pads with over a couple of hundred wells. The study was focused on frac-hit interaction among paired wells and demonstrated that the ML model, once trained, can predict the frac-hit probability.
This study implemented a data-intensive supervised machine learning approach, through a quasi-experimental framework, with the objective of quantifying the impact of oil and gas operator-specific proprietary data on ML-based predictive model performance as compared to using oil and gas datasets that are publicly available.
This work was focused on developing novel methods to facilitate rapid data to knowledge transformation for subsurface applications, particularly those related to unconventional oil and gas.
Here a convoluted process model-filtering technique is presented that can build and successfully train the structural property artifacts of materials after multiple heat treatment cycles.
In this work, compute acceleration of such model training was achieved by taking advantage of extreme workflow parallelism of single-wafer-embedding of the compute resources.
In this presentation, ‘deep-freeze’ graphs, ‘convoluted filtering’ networks, ‘mirror-image’ graphs, and adversarial ensemble methods are utilized to support inversion modeling for optimization of the complex compositions and complex processes in design of high-performing alloys, with their properties tailored to the energy application specifications.
Principles of materials science and engineering, physics, mathematics, and information science are used to extract knowledge and insights from the process-structure–property-performance relationships hidden in materials data. The process-structure modeling can be accelerated without loss of interpretability, with artificial intelligence tools that mimic the salient features of the process and process-structure relations. In this work, a novel convoluted model-filtering technique was exploited to build and successfully train the Convoluted Filter (CoFi) artifacts for Fe-based alloy heat treatment cycles. The artifacts were pre-trained to filter out deep models that change the surrogate microstructure state after the heat treatment at ambient conditions. Direct representation of the thermal cycle features within knowledge Graph facilitated development of meaningful data models for microstructure evolution, which reduce overfitting to limited datasets.
This study focusses on the development of an ML-based workflow that can be used for frac-hit identification and monitoring based on the readily available field data, such as drilling surveys, pressure records, production history (oil and water production measurements), and fracking time and duration information, etc.
International Meeting for Applied Geoscience & Energy (IMAGE '22), Houston, TX, August 28-September 1, 2022
XMAT Workshop 2022, Virtual, March 11, 2022
This paper is for the Proceedings of the IEEE Big Data 2021 Conference (IEEE BigData 2021) published by IEEE Computer Society.
<span style="font-family: Calibri, sans-serif; font-size: 12pt;">Accelerated materials design should match the recent trends in the product development cycles. Materials data analytics can be used to significantly shorten development time of specialized alloys needed for next generation energy applications. However, it faces a challenge of scarce data available for training ML models. Incorporation of the domain knowledge into deep-learning graph structure via fuzzy pre-training and causal process imitation presents a viable approach to developing accurate data-driven models and reliable alloy design tools, with limited datasets. Artificial Intelligence (AI) was used in this study to incorporate such knowledge in the domain-specific computational tool, pyroMind. The tool provides not only novel design ideas but also their interpretation via physics and engineering concepts.</span>
<span style="font-family: Calibri, sans-serif; font-size: 12pt;">Materials data analytics can be used to significantly shorten development time of specialized alloys needed for next generation energy applications. Incorporation of the domain knowledge into deep-learning graph structure via fuzzy pre-training and causal process imitation presents a viable approach to developing accurate data-driven models and reliable alloy design tools, with limited datasets. It was demonstrated that the domain knowledge-based empirical constraints not only inhibit overfitting but also allow training more accurate and reliable ML models with improved transparency of the output interpretation. In this study, alloy tensile properties were interpreted with three competing virtual-microstructure models.</span>
<span style="font-family: Calibri, sans-serif; font-size: 12pt;">Continuous improvement in efficiency of a power plant relies on designing materials for use at increasingly higher temperature and/or pressure, for 100,000s hours of operation. Due to complexity, non-linearity and high-dimensionality of the problem, traditional Machine Learning (ML) approaches require unreasonably large datasets for the data-driven model development. Science-based material and process engineering complements hard data with, sometimes soft and intuitive, empirical domain knowledge. Artificial Intelligence (AI) was used in this study to incorporate such knowledge into computational graph architecture (process-mimicking artificial neuron design, causal layer and graph structures, ensemble modeling of latent states) and learning procedures (variable transformation, fuzzy physics pre-training and freezing of deep layers, virtual microstructure representation, and adversarial multi-objective optimization). The first alloys design pathways suggested by the AI tool (pyroMind) passed a preliminary engineering review on soundness and transparency.</span>