desulphurisation

There are certain grades which have a strict limitation on max-Si permitted, typically 0.03% (wt), so it’s necessary to avoid silicon reversion from slag (that happens when we “OVER” kill slag[FeO] with Al).

            Normally de-O precedes de-S. Once SUFFICIENT de-O is achieved, de-S should be initiated WITHOUT delay & deep de-O should be avoided. In order to de-O, we add Al in ONLY-Al killed grades. Once added, this aluminium reduces FeO(unstable, weak oxide) first. THEN…., it looks for other oxides to reduce, when this heavily-reducing atmosphere is maintained unbalanced. Once free oxygen and FeO are killed, now Al looks at silica and MnO, and starts reducing them- unless CaO (lime) is added to raise the basicity.

Once lime is added, reaction-collisions (Al + SiO₂ =>Al₂O₃+ Siˇ) decrease, leading to lesser Si reversion. aluminium is now surrounded by more CaO molecules than SiO₂ molecules, so Al now attacks CaO and develops Ca-ions. These Ca-ions attack FeS and START de-S reaction. So instead of [Al + SiO₂ =>Al₂O₃+Siˇ), [Al + CaO => Ca⁺ + Al₂O₃ ] reaction dominates, enabling better de-S.

So when it comes to Al-killed grade, we need to control slag very carefully. De-O of only FeO & MnO should be permitted. De-O of silica should NOT be allowed to happen.

That means…in Si-killed grades, slag can be maintained fluidic FOR PROLONGED DURATION, while in Al-killed grades longer durations of reducing atmosphere+fludic slag returns Si into metal. Hence CaO addition at the RIGHT time* in RIGHT quantity^$ is very important. Excess fluidity is not permitted for longer duration.

RIGHT time*- because late addition, would have already resulted in Si reversion into metal.

RIGHT time*- because early addition, will impede de-O RATE.

RIGHT quantity^$ - because too much, will reduce de-S RATE due to reduced reaction-collisions.

RIGHT quantity^$ - because too less, will permit Si-reversion & will reduce de-S as slag becomes easily S-saturated.

We can NOT add lime in excess quantity (basicity should not go beyond 3.2). Excess lime will make slag very thick/dry/stiff, so de-S RATE will be heavily reduced due to reduced reaction-collisions.

             This thick/dry/stiff slag may lead to arc-flare as arc gets deflected, which may reduce ladle-lining-life, apart from reduced de-S rate.

            And, thin/fluidic slag will NOT bury the arc, leads to energy waste, arc flare may also happen. ULC/IF grades will pickup too much carbon from electrode. May lead to electrode spalling.

            So we have to walk a thin-line between too-thin slag and too-thick slag in only-Al-killed, de-S grades, where Si reversion is NOT permitted (Si < 0.03 wt%).

            In Si-killed grades de-S becomes easier, since Si-reversion is permitted; delayed CaO addition can be done in too many multiple batches of very small quantities.This permits slower thickening of slag, thereby expediting de-S through increased reaction-collisions of fluid slag molecules.

            Finally, one more point I shall add. It is not Al or Si that decides the basicity but the oxides of them. Both oxides are acidic slag formers, making slag more fluidic.

About amphoteric nature of alumina in basic-steel-making practices, I have no clear idea; though I may venture to say, aumina acts as an acidic slag in basic steel making conditions.