Establishment of Thermodynamic Properties Datasets of Binary Molten Chloride Salts for Concentrating Solar Power Applications: Using Machine Learning Strategies

Project: Research

Project Details

Description

Concentrating solar power (CSP) plants with thermal energy storage systems are capable of generating continuous electricity, but their performance largely depends on the thermal properties of the heat transfer fluids. The molten binary chloride salts are a promising candidate for the heat transfer fluid in Gen 3 CSP due to their low cost and stability, but their thermal properties need to be determined accurately. Despite a wealth of data reported in the literature, the reliability of these results is questionable due to experimental difficulties. Molecular simulation schemes based on first principles also face obstacles in accurately describing the molecular interactions of chloride salts.To address these challenges, this project aims to establish reliable thermal property datasets for binary alkali chloride salts using machine learning strategies. Machine learning with molecular dynamics (MD) will be used to construct potential functions for the ion pairs in the melts of interest, enabling direct simulations. An active learning framework will be used to compare and extract countable literature results, with herein yielded numerical results. The two machine-learning schemes will be applied to establish a platform to set up the thermodynamic properties of molten binary chloride salts for the Gen 3 CSP applications. This project will accomplish the following:(1)The existing VASP-DeePMD-kit scheme will be applied to establish the potential functions of ion pairs of specific chloride salts and their binary mixtures (NaCl-LiCl, NaCl- KCl, NaCl-BaCl2) over Gen 3 CSP working conditions.(2)The active learning platform built by Python will be established to extract the thermodynamic properties and nanostructure information of molten salts (LiCl, NaCl, KCl, MgCl2, CaCl2, ZnCl2, BaCl2, and their mixtures) from the available literature or the present simulation platform.(3)The datasets for thermodynamic properties (density, isobaric heat capacity, thermal conductivity, viscosity) of the molten binary chloride salts (NaCl-LiCl, NaCl-KCl, NaCl- BaCl2) over 0-100% w/w and 800-1000 oC, atmospheric pressure will be established. Novel datasets with SiO2 nanoparticles and under an external electric field for these molten salt mixtures will also be established.The success of this project will contribute significantly to the development of Gen 3 CSP technologies. The long-term impact of the project includes facilitating the development of advanced heat storage technology that the field application of renewable energy urgently needs and demonstrating how artificial intelligence technology can be applied to design novel molten salt heat transfer and storage materials effectively. 
Project number9043660
Grant typeGRF
StatusActive
Effective start/end date1/01/25 → …

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.