Mineral Particle Deposition in High Temperature Processes

Global trends such as resource scarcity and climate change due to industrialization and urbanization, have attracted growing attention from researchers. To reduce CO₂ emissions, a transition from conventional fuels e.g. coal and coke, to alternative fuels e.g. natural gas and solid recovered fuel (SRF), is taking place in stone wool (ROCKWOOL) and cement (FLSmidth) production plants. However, this transition is challenged by the formation of mineral deposits during the production processes, which is induced by the use of alternative fuels and/or the introduction of new production technologies. The excessive deposits formed by ash and/or mineral raw materials usually lead to several operational issues, such as reduced heat transfer efficiency, variations of product quality, blockage of flue gas channel or even unscheduled plant shutdown. Therefore, reducing unwanted mineral deposit formation and timely deposit removal are essential for the optimal production of stone wool and cement. In this PhD project, an improved understanding of the mineral deposits formation, sintering, and removal has been obtained through experimental work in laboratory-scale setups. Furthermore, the characterization of industrial deposit samples from cement plants has been performed. Finally, a mechanistic model was developed to describe the mineral particle deposition behavior in an entrained flow reactor with stone wool raw materials.

Deposition experiments using raw materials for stone wool production were conducted in a lab-scale entrained flow reactor (EFR). To mimic the temperature conditions of the stone wool production process, a ceramic deposit collecting probe system which is inserted into the reactor via an elevator was developed and employed. The effects of fed materials (different raw materials: Anorthosite, Serox, Merox, Diabase, Dunite, Wool waste; and different charge mixtures: charge1 and charge2), furnace temperature (700℃-1200℃), percentage of Wool waste in charges (0%-30%), percentage of Serox in charge1 (0%-20%), gas velocity (0.9m/s-1.8m/s), and particle size distribution on deposit formation rate were investigated. The primary mechanism of mineral particle deposition was deduced as follows: high temperature melts/softens part of the raw materials, causing them stick to the ceramic probe via inertial impaction. Only Diabase, Wool waste, and two charges formed a considerable amount of deposits at a furnace temperature of 1100℃, which was qualitatively correlated to the deformation temperature and Si/Al ratio of applied materials. The deposit formation rate of charge2 remained at a low level (<200g/(m²*h)) from 700°C to 1000°C, however, a significant increase was observed at 1100°C, being 25000g/(m²*h). Furthermore, increasing the percentage of Wool waste in the two charges significantly increased the deposit formation rate, because the melting/softening of wool waste would increase the stickiness of the deposit surface thereby increase the capture efficiency of the impacting particles. However, increasing the gas velocity and particle size had a negative impact on the deposit formation rate, due to the increased rebound probability and less melting/softening degree of impacted particles. Additionally, the practical implications of industrial stone wool production were discussed. It is found that the most important characteristic of deposit formation rate is the weight percentage of sticky materials (Wool waste > Diabase), followed by the Si/Al ratio, which indicates the material deformation temperature.

Furthermore, the experimental investigation of cement mineral particles deposition behavior was conducted in a laboratory-scale entrained flow reactor, as well. To simulate the temperature conditions in a cement calciner, two different deposit probe systems were used: a high probe surface temperature (HPST) deposit system with a probe surface temperature of 700°C to 1200°C and a low probe surface temperature (LPST) deposit system with a probe surface temperature of 500°C-700°C. The effects of fed materials (raw meal, hot meal, bypass dust), flue gas temperature (700°C-1200°C), deposit probe surface temperature (550°C-1200°C), gas velocity (0.9m/s-2.67m/s), and experimental duration (5min-60min) on deposit formation rate were investigated. The results revealed that the concentration of KCl in the fed materials has a large influence on deposit formation. In the HPST experiments with furnace temperatures ranging from 700°C to 1100°C, the bypass dust has a higher deposition rate than hot meal and raw meal, due to a large amount of K and Cl (>10% wt%) in the bypass dust, which forms melts and increases the stickiness between the impacting particles and deposit. However, in LPST experiments, the presence of KCl facilitated deposition rate by providing a sticky layer on the deposit probe via condensation, resulting in a higher deposition rate of bypass dust than raw meal. The gas velocity showed a negative effect on the deposit formation rate due to an increase of rebound. Finally, the practical relevance of deposits formation in calciner was discussed. It was observed that intermediate temperatures (900 to 1100℃) and keeping the K content at a low level are beneficial to the reduction of deposit formation.

The characterization of industrial deposit samples is helpful to understand the mechanism of full-scale mineral deposit formation processes. Six industrial deposit samples from a cement plant co-firing coal and solid recovered fuel (SRF) were analyzed both qualitatively and quantitatively by QRXD, SEM-EDS, ICP-OES, and STA. The results revealed that KCl, CaCO3, and spurrite (Ca₅(SiO₄)₂CO₃) are the most prevalent crystal phases in all the deposit samples. The chemical compositions of analyzed samples differed greatly between different locations, which is attributed to the difference in local temperatures. The deposit formation was found to be promoted by KCl condensation on the cold surfaces of equipment walls and particles, forming a sticky layer for subsequent deposits. According to the STA results, the decomposition temperature of spurrite was inferred to be around 900°C – 1080°C. Furthermore, the comparison of chemical composition between full-scale deposit samples and experimental deposit samples was performed, which shows a significant difference in quantities of K, Cl, Ca, and Si. In addition, two mechanistic models explaining the mineral particle deposit formation processes in different local temperature areas were proposed.

A study on the adhesion strength of mineral deposits was conducted on a mixture of KCl and cement mineral materials. In this context, the adhesion strength of artificial mineral deposits prepared with cement raw materials and KCl was quantified in a lab-scale force gauge oven to obtain an improved understanding of mineral deposits removal. The effects of materials (raw meal, hot meal, and bypass dust), sintering temperature (700℃ – 850℃), percentage of KCl in the mixture (0 wt% – 60 wt%), measurement temperature (700℃, 750℃, and 800℃), and sintering time (4h – 24h) on adhesion strength were studied. The obtained findings revealed that the increase in adhesion strength is mainly due to the melting and solidification of KCl particles in the model deposits. The adhesion strength remained low, around 3 kPa, as the sintering temperature increased from 700°C to 850°C, with the exception of a sharp increase at 750°C attributed to the partial melting of KCl particles. Furthermore, as the amount of KCl in the mixture increased from 0 wt% to 60 wt%, the adhesion strength increased considerably from 3 kPa to 440 kPa at a sintering temperature of 750℃. A lower measurement temperature of a pre-melted deposit increased the adhesion strength which is explained by the solidification of melted KCl during cooling. However, because of the low KCl concentrations, different pure cement materials had little effect on adhesion strength.

Finally, a mineral particle deposition model was employed to describe mineral particle deposition in an entrained flow reactor with stone wool raw materials. The model accounts for the inertial impaction deposition mechanism and heating up of particles via convection and radiation. The critical velocity sticking probability submodel was employed to describe the sticking behavior of the impacting particle. The deposit formation rates at various furnace temperatures (700℃ – 1100℃) and gas velocities (0.9 – 1.8m/s) with charge2 were simulated. The findings showed that the overall simulated results at different furnace temperatures and gas velocities of 0.9m/s and 1.2m/s agreed with the experimental data with the application of a new correlation of Young’s modulus fitted to the experimental data of charge2. However, the modeled deposition rates at higher gas velocities (1.5m/s and 1.8m/s) showed relatively large deviations which is probably due to the assumption of uniform plug flow and negligence of particle agglomeration. Furthermore, the improvement of the current model was suggested in two respects, i.e. the development of an impaction model based on an inclined flat surface and the precise correlation of Young's modulus of applied materials and temperature.

To conclude, this work provides an improved understanding of the mineral particle deposition in high temperature processes. The observed trends obtained in this work show the potential for improvement of design and operation to minimize the unwanted mineral particle deposition during stone wool and cement production.