Utilizing low-rank coals in power plants takes a risk of slagging, especially when burning coals rich in alkali and alkaline earth metallic (AAEM) elements. To predict the slagging propensity of low-rank coals, this work develops a novel approach for accurately measuring the sticking probability of thermally stable slag samples based on the well-controlled flat-flame system. We collect deposit samples (PSD) severely accumulated on the platen superheater surface of a 660 MW boiler co-firing two low ranks (alkali-rich Phi-coal: iron-rich Ind-coal = 15.2 wt: 84.8 wt). It is found that the existing theoretical models can hardly reproduce the measured sticking probability at elevated temperatures. We thus propose an Sshaped nonlinear correlation of ash sticking probability with the melt fraction using two tuning parameters. The fully-sintered PSD slag is mainly composed of refractory anorthite (CaAl 2 Si 2 O 8 ) and hematite (Fe 2 O 3 ), which are largely determined by the dominant Ind-coal ash fed into the boiler. However, the excessive accumulation of PSD can only be attributed to the much stickier Phi-coal ash rich in lowmelting-point, dispersed Na-aluminosilicates. Combining the measured sticking probability and physicochemical properties of the slag and ash samples, we reveal a positive correlation of the ash sticking probability with the sodium content of the samples, almost regardless of the contents of alkaline earth metals and iron. It highlights the leading role alkali metals play in slagging. In contrast, iron exists as discrete hematite (Fe 2 O 3 ) in the slag and is less effective than AAEM in forming slags on the platen superheater. Our work provides a more reliable method and criterion for evaluating the slagging propensity of fully/partially-molten particles in solid fuel combustion. (c) 2021 The Combustion Institute. Published by Elsevier Inc. All rights reserved.
展开▼
