Recent Advances in High-Fidelity Simulations of Pulverized Coal Combustion
Ruipeng Cai,Kun Luo,Hiroaki Watanabe,Ryoichi Kurose,Jianren Fan
DOI: https://doi.org/10.1016/j.apt.2020.05.001
IF: 4.969
2020-01-01
Advanced Powder Technology
Abstract:Although renewable energy has recently attracted a lot of attention, coal-fired power plants will still play an important role in electricity production in the future due to coal's abundance in the world. However, combustion of pulverized coal is extremely complex which makes experimental measurements very challenging even impossible. As an alternative, computational fluid dynamics (CFD) is a powerful tool to investigate pulverized coal combustion (PCC), and can help to design and develop advanced coal-fired facilities. With the development of computational capacity, high-fidelity simulations of PCC have been developed and significant achievements have been obtained in the recent years. In this review, the recent advances in CFD of PCC, especially in fully-resolved direct numerical simulation (DNS), point-particle DNS and large eddy simulation (LES) of PCC are summarized. The progresses on both numerical methods and coal combustion models are highlighted. The state-of-the-art applications of these high-fidelity simulations are also reviewed, including the investigation of combustion characteristics of pulverized coals, the development of advanced coal combustion models, and the development of clean coal technologies. It is concluded that fully-resolved DNS, point-particle DNS, and LES have their merits and demerits, so the optimal use of each approach depends on specific research purpose. For coal devolatilization and char-oxidation models, the combined models in which advanced models are used to calibrate the kinetic parameters of simple models are quite attractive in terms of computational accuracy and computational burden. For chemical reactions, the flamelet model is popular, but the predictive capability and robustness need to be further improved. Besides, LES is rapidly approaching to simulate industrial PCC. However, the computational cost is still quite expensive, and more efficient algorithms and sub-grid models are required. To this end, DNS database, benchmark measurements, and international collaborations are important. Machine learning may also play an important role in promoting the development of high-fidelity simulations of PCC in the future. (C) 2020 The Society of Powder Technology Japan. Published by Elsevier B.V. and The Society of Powder Technology Japan. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).