The bottom blown copper smelting furnace is a novel technology developed by Dongying Fangyuan Nonferrous Metals Co. Ltd. China. Since it started in 2008, many advantages have been reported from industrial practice, such as high volume specific smelting capacity, low copper loss, autogenous smelting, adaptive to feeds, low temperature operation and so on. Most of those features are relevant to the fluid dynamics of the molten bath in the furnace. However, the detailed behaviour of molten bath in the bottom blown reactor still is not comprehensively understood.
To understand the behaviour of molten bath in this newly established furnace, a 1/12 lab scale cold model was constructed according to the principles of similarity. In this physical model, water was used to simulate the liquid matte, compressed air was used to simulate oxygen enriched air blowing and its flowrate was adjusted accordingly. Silicone oils with different viscosities were used to simulate the molten slag layer of various viscosities. In the present study, several features of the bottom blown bath behaviour were investigated by using this physical model.
The mixing behaviour in the single phase bath with single lance blowing was investigated in the first stage. Mixing time was used as the index of mass transfer efficiency in this cold model and was measured to examine the characteristics of the fluid dynamics in the bath. It was found that there is an effective stirring range within an area adjacent to the blowing lance, in which mixing time changes little with horizontal or vertical distance, while out of the range mixing time increases with increasing horizontal distance and the increment is much greater on the surface of the bath. Mixing time was found to decrease with increasing bath height and gas flowrate within the effective stirring range. However, the effective stirring range can be reduced with increasing the bath height. An empirical prediction of mixing time from gas flowrate and bath height was hen established for horizontal bottom blown furnace with single phase bath.
[See pdf for formula]
In addition to single phase mixing behaviour study, multiphase bath mixing behaviour was also studied to further investigate the mass transfer in the presence of top layer. Single lance blowing was used and experimental variables including water height, gas flowrate, oil layer height and oil viscosity were adjusted to investigate the impact of each on mixing. It was found that the mixing time decreases with water height and gas flowrate in a greater rate than that of single phase system, while it is increasing with oil layer height and oil viscosity, where the rate with oil layer height is much greater. An overall empirical relationship with these variables was developed as following, and it showed a good prediction of mixing time under different conditions.
[See pdf for formula]
The correlation was then generalised to a model-independent format for wider use:
[See pdf for formula]
Three different types of surface waves were studied in the single phase bath with single lance blowing. The 1st asymmetric standing wave was found the most significant type among the three because it leads the entire bath to swing laterally. The 1st asymmetric standing wave was found to take place only at certain combination of bath height and gas flowrate, which is summarised in two sub-boundaries in terms of bath depth, gas flowrate and blowing lance angle.
Sub-boundary 1 [See pdf for formula]
Sub-boundary 2 [See pdf for formula]
The measured amplitudes at several conditions show that tapping end amplitude increases with flowrate and bath height, and when the 1st symmetric standing wave is taking place, the amplitude is remarkably increased. The frequency of the 1st asymmetric standing wave does not change with flowrate or blowing angle, but slightly increases with bath height.
The longitudinal wave exists on bath surface when the combination of bath height and gas flowrate is out of the critical occurrence boundary. To study the behaviour of the longitudinal wave, the model dynamic condition was constructed closer to industrial practice. Nine lances were used for blowing and silicone oil was applied on water surface to investigate the behaviour of longitudinal wave. It was found that the amplitude trend of longitudinal wave along the furnace length is dependent on the combination of gas flowrate and bath level, and three typical trends were observed.
At the gas flowrate corresponding to industrial flowrate, the amplitude of wave increases with water level, but remains almost stable with oil level. Higher viscosity oil layer leads to lower amplitude of wave, while lower viscosity oil leads to rising of amplitude near tapping end. The amplitude of wave at the edge of tapping end is found always higher than the centre. The frequency of longitudinal wave is not affected by the water level, oil level or oil viscosity.
All these features are linked to the industrial operation of this novel copper smelting furnace. The investigation in this study provides guidelines and useful information for understanding and further improvement of bottom blown copper smelting technology in the future.