APPLICATIONS OF MACHINE LEARNING ALGORITHMS IN INTEGRATED CARBON CAPTURE-MINERALIZATION (ICCM) PROCESSES

Other Titles
Abstract

Regulation of CO2 emissions is imperative to prevent environmental catastrophes brought on by climate change. Technologies for carbon capture and storage are being developed extensively as environmentally friendly solutions to the issue. Traditional methods divide carbon capture and storage into two phases, leading to inefficient operation and excessive energy usage. To expedite the reactions and enhance productivity, an integrated carbon capture_mineralization process was proposed at a later stage. The classic carbon-water-mineralization system is improved by this combined carbon capture and mineralization technology by removing the constraint of limited carbon dioxide solubility. However, aside from the advantages, the integrated carbon capture_mineralization processes suffer from some limitations, such as instability of amine-based solvents or gel-like formation during the reactions. In addition, competing phenomena exist among amines, such as tradeoffs between reaction rates and viscocities, so it is insufficient to examine a single property of the amine for improving mineralization performance. Thus, for researchers and engineers to increase mineralization efficiency, simulations were used to help optimize process schemes or selecting the best reactants by comprehensive analysis on physical, chemical, and reactions properties of amines involved in carbon capture and mineralization processes. Aspen Plus was used to simulate the two-stage integrated carbon capture_mineralization processes and to __identify reinforcing feedback between mineralization and related parameters. In the simulation, CO2 was first removed from the flue gas stream using amine-based solvents (monoethanolamine diethanolamine, triethylamine, and blended monoethanolamine+Methyl diethanolamine). After this, an alkaline substance (CaO) was used to regenerate the CO2-loaded amines through mineralization processes, and the final desorbed CO2 was transformed into bicarbonates, which were then mixed with calcium ions (Ca2+) to create calcium carbonate (CaCO3). The Aspen-generated data in the simulation processes were then used for the next part: implementing machine learning algorithms for enhancing CO2 mineralization level in the integrated absorption_mineralization processes. Machine learning algorithms use statistical calculations to predict outcomes and find optimizations. The objective of this study is to forecast how different amines will perform on carbon mineralization under various reaction conditions by using well-know supervised machine learning algorithms, such as Artificial Neural Networks (ANN), Support Vector Regression Machine (SVM), K-nearest neighbor (KNN), Decision Tree Regression (DTR), Lasso Regression, and Ridge Regression(RR), Using multiple statistical indicators, our findings show that KNN outperformed other models in terms of Root Mean Square Error (RMSE) (0.027), Mean Absolute Error (MAE) (0.012), Mean Absolute Percentage Error (MAPE) (0.479%), and correlation coefficient (R2) (0.9929) values. The accurate predictive modeling is of great importance for potential industrial advancement and solving CO2 emission problems, as it can find optimal reaction conditions. The combined “machine learning-integrated absorption and mineralization” is beneficial for the development of smart carbon capture and storage technologies and the creation of efficient reaction systems, while bridging the gap between artificial intelligence and environmental protection.

Journal / Series
Volume & Issue
Description
88 pages
Sponsorship
Date Issued
2022-12
Publisher
Keywords
amine solvents; Aspen Plus; integrated carbon capture and mineralization; Machine Learning; optimization; post-combustion
Location
Effective Date
Expiration Date
Sector
Employer
Union
Union Local
NAICS
Number of Workers
Committee Chair
Gadikota, Greeshma
Committee Co-Chair
Committee Member
Helbling, Damian
Barstow, Buz
Degree Discipline
Civil and Environmental Engineering
Degree Name
M.S., Civil and Environmental Engineering
Degree Level
Master of Science
Related Version
Related DOI
Related To
Related Part
Based on Related Item
Has Other Format(s)
Part of Related Item
Related To
Related Publication(s)
Link(s) to Related Publication(s)
References
Link(s) to Reference(s)
Previously Published As
Government Document
ISBN
ISMN
ISSN
Other Identifiers
Rights
Attribution 4.0 International
Types
dissertation or thesis
Accessibility Feature
Accessibility Hazard
Accessibility Summary
Link(s) to Catalog Record