The main objective of this work was to gain an increasedunderstanding of selected blast furnace phenomena which couldbe utilized for an improved blast furnace process control. Thisthesis contributes with both a model study and an experimentalstudy on blast furnace tapping, and results from these findingscan be used to enhance the control of the blast furnace.
The work was divided in two parts. The first part dealt witha model study for optimisation of the blast furnace burdencalculation. During the second part the frequency of thehot-metal and slag sampling was increased compared to routinesampling throughout the taps of a commercial blast furnace.Thereafter, composition variation and correlation betweendistribution coefficients were examined.
With an optimisation of the burden calculation the firststep towards controlled hot-metal production is taken, sincethe optimal material mixture for a desired hot-metalcomposition could easily be found. Due to the fact that theoptimisation model uses yield factors, which are easy tocalculate from material and hot-metal compositions, thesevalues have to be accurate for a controlled process control ofthe furnace. The study of hot-metal and slag compositionsduring tapping concluded that variations exist. The largevariations for C, Si, S, Mn and V in hot metal during tappinglead to the conclusion, that one single sampling ofhot metalwas not enough to get a representative value for thecomposition. The solution was to use a double-samplingpractise, were the hot metal was sampled first after tap startand secondly short after slag start, and subsequently anaverage composition value was calculated. The following studywas on the elemental distribution between hot metal and slagfrom a thermodynamic point of view. The major conclusion fromthis study was that the distribution coefficients behaved asexpected when looking at the equilibrium reactions. The studiedslag-metal distributions were also showing strong, trend-likerelationships, which was not affected by the operational statusof the blast furnace during the studied sampling period.
The overall conclusion is that with a more reliablecomposition of hot metal and slag from the taps, thedistribution coefficients could be calculated with betterprecision and hence, the yield factors for the optimisationmodel would be more accurate. This procedure would probablylead to a more reliable burden optimisation and a thereforebetter and more stable blast furnace control.
Stockholm: Materialvetenskap , 2003. , x, 25 p.