Electric-arc furnace dephosphorisation with high-phosphorus DRI: is the target achievable?
An electric arc furnace taps 120 t of steel from a charge of 84 t high-phosphorus DRI (0.12% P, 3.5% SiO2) and 36 t low-phosphorus scrap (0.03% P). The customer requires 0.020% P at tap for sour service. Current practice is single-slag: lime to a slag basicity CaO/SiO2 of 2.0, a slag weight estimated at 4.0 t, 18% FeO at turndown, and tapping at 1,640 °C with 0.05% C. Turndown analysis gives 0.045% P. The shop has no slag-raking station but does have eccentric bottom tapping and a ladle furnace. Lime is plentiful; oxygen and tap-to-tap time are the constraints. The melter reports that slag foaming is poor and the slag looks stiff at turndown.
1. Carry out a phosphorus balance and calculate the phosphorus partition ratio L_P = (%P2O5)/[%P] needed to hit 0.020% P versus what the current practice achieves. What does the gap tell you?
Read the worked answer
Phosphorus in: 84 × 0.0012 = 0.1008 t plus 36 × 0.0003 = 0.0108 t, so 0.1116 t = 111.6 kg P.
Target: 0.020% of 120 t = 24 kg P in steel, so 87.6 kg P must report to slag. As P2O5: 87.6 × 142/62 = 200.6 kg. Over a 4.0 t slag that is 5.0% P2O5, giving L_P = 5.0/0.020 = 250.
Present: 0.045% P in steel = 54 kg, so 57.6 kg P in slag = 132 kg P2O5 = 3.30% P2O5, giving L_P = 3.30/0.045 = 73.
Interpretation: the required partition ratio is 3.4 times what is being achieved. L_P cannot be raised from 73 to 250 on the same slag by a small lime addition - the sensitivity of L_P to basicity and FeO is not that steep. Note also that if L_P stayed at 73, holding 0.020% P would need 1.46% P2O5 in slag, i.e. 87.6 × 2.29/0.0146 ≈ 13.7 t of slag, which the furnace cannot carry.
The numbers also expose an inconsistency worth flagging: 84 t of DRI at 3.5% SiO2 brings 2.94 t SiO2. Basicity 2.0 on that alone needs 5.9 t CaO, so a 4.0 t slag with basicity 2.0 is impossible. Either slag weight or basicity is mis-reported and slag is being carried over at tap. Fix the measurement before trusting any L_P number.
2. Recommend a practice change and explain the process conflicts you must manage. Support the recommendation with the quantities from the balance.
Read the worked answer
Recommended route - double slag with an early high-FeO flush:
- Melt down and blow to about 1,600–1,620 °C with basicity 3.0–3.5, FeO held at 22–28%, and a slag weight of 10–12 t (enough lime to cover 2.94 t SiO2 at basicity 3). At 1,620 °C rather than 1,640 °C, L_P is materially higher; with basicity ~3 and FeO ~25%, L_P of 150–250 is realistic for this carbon level.
- Deslag fully, then rebuild a small reducing slag for desulphurisation and alloying in the ladle furnace, and tap slag-free by eccentric bottom tapping.
- Do not attempt the whole duty in one slag while also finishing carbon: the furnace cannot simultaneously hold 0.05% C and a 5% P2O5 slag economically.
Conflicts to manage:
- Temperature. Dephosphorisation wants low temperature, but the slag must stay fluid and the tap must be hot enough for the ladle furnace. The flush at 1,600–1,620 °C then a controlled reheat resolves this.
- FeO versus yield and deoxidation cost. High FeO partitions phosphorus but burns iron and demands more deoxidiser. Take the FeO penalty early and drop it before tap.
- Basicity versus slag volume and foaming. More lime means more slag, higher viscosity, and the stiff slag already reported. Improve fluidity with fluorspar or a flux, not by dropping basicity.
- Carbon. Oxygen that goes to carbon is not available to hold FeO up; finish the carbon after the phosphorus flush.
3. Why does tap temperature and slag carryover control whether the phosphorus stays in the steel? Explain the reversion risk during tapping and deoxidation.
Read the worked answer
The phosphorus in slag is held as 2CaO·SiO2–3CaO·P2O5 (or as a phosphate in the slag's basic network). That hold is an equilibrium with the metal: L_P falls as temperature rises and as the slag's FeO and basicity fall. Two things happen after the turndown sample is taken.
During tap: the metal is hotter than at turndown if the furnace reheats, and high-FeO slag is entrained into the ladle by the tapping stream. Entrained slag is a small fraction of the furnace slag but it is in intimate contact with 120 t of metal that has a very large capacity to reduce P2O5 once its oxygen potential drops.
During deoxidation and alloying: adding aluminium, ferrosilicon and ferromanganese drops the dissolved oxygen and the slag FeO/MnO sharply. The reaction P2O5 (slag) + 5[Fe] → 2[P] + 5(FeO) is driven left by FeO and right by deoxidation, so phosphorus reverts from any carried-over slag straight back into the metal. A 0.020% P heat can finish at 0.030% P after a heavy aluminium addition if the ladle has a thick oxidising slag cover.
Controls: tap as cool as the ladle furnace allows (below about 1,630 °C), use eccentric bottom tapping or a slag stopper to keep furnace slag out of the ladle, keep the ladle slag small and either reducing or deliberately removed, and sample phosphorus on the ladle furnace after deoxidation rather than trusting the turndown number.