A report on Kinetic Modelling of External Desulphurization by Powder Injection

A report on Kinetic Modelling of External Desulphurization by Powder Injection

1. A Report on Kinetic modelling of external desulphurization by powder injection Vaibhav mishra Dr.Rahul SARKAR

2. Contents Theoretical area required for desulphurization Powder injection process Major Literature works on kinetic modelling Approaches for kinetic modelling Work plan Reaction zones finalization Sample simulation and its validation Reagent optimization study

3. Total Interfacial Area Required Sulphur target(ppm) 5 Sulphur at equilibrium(ppm) 2 Sulphur initial(ppm) 6 00 Ratio 0.051724 Diameter of ladle(m) 4 Metal bath depth(m) 3.70 Mass of metal(kg) 288000 Volume of metal( m 3 ) 46.472 time(sec) 900 km(m/sec) 0.00001 s1 -2.96183 A( m 2 ) 15294 Area of ladle( m 2 )(top slag –metal interface) 12.56 Related equation: e / i - e )= p k m t)/V m [%S]= wt.% sulphur at time t [%S] e = wt.% sulphur at equilibrium [%S] i =wt.% sulphur initially in HM A p = projected area required(m 2 ) k m = mass transfer coefficient(m/sec) V m = volume of HM(m 3 ) t = time(sec)  

4. Mg+CaO co-injected with a carrier gas Major Reaction zones are: a.Transitory Reaction zone b.Permanent Reaction zone Transitory reaction zone Bubble plume (HM+ flux particles +Gas bubbles) Permamant reaction zone Top slag and HM 4 Powder Injection Process Overview Mg+CaO+N 2 gas Submerged lance Transitory Reaction Zone Permanent Contact Reaction zone Off-Gas Elmira Moosavi- Khoonsari, M.-A. Van Ende and I.-H. Jung: Metall. Mater. Trans. B., 2022, vol. 53 B, pp. 981-998 WENTAO LOU and MIAOYONG ZHU : Metall. Mater. Trans. B., 2014, vol. 45B, pp. 1706 -22 Bubble plume Hot Metal Top slag

5. Main Reactions in Transitory and Permanent contact zones Transitory reaction zone Permanent reaction zone (Dissolved Magnesium and sulphur reaction in plume) (Lime particle and dissolved sulphur reaction in plume)   Absorption of & with unreacted to top slag   , & + bubble plume   2 )   Skimming of slag 5 Elmira Moosavi- Khoonsari, M.-A. Van Ende and I.-H. Jung: Metall. Mater. Trans. B ., 2022, vol. 53 B, pp. 981-998 H.-J. Visser: Modeling of Injection Processes in Ladle Metallurgy, Ph.D. Thesis, 2016, TU Delft, the Netherlands

6. Approaches available for Kinetic Modelling Metal Slag Reaction Interface Modelling Distance Concentration A AO k m Equilibrium at interface = ( -   = (stirring energy, diffusivity)   Real process Very Complex Individual mass transfer equations Difficult to find mass transfer coefficients k m & k s Changes by process conditions Mass transfer Diffusion boundary k s 6 M.-A. Van Ende and I.-H. Jung: Metall. Mater. Trans. B ., 2017,vol. 48B, pp. 28–36

7. Simplified overall-mass transfer approach Easy to apply to complex process Easy linkage to FACTSAGE database Approaches available for Kinetic Modelling Metal Slag V 1 V 2 V 3 V 4 Equilibrium in reaction zone volume Reaction amount at interface: Metal = metal Δ t Reaction amount at interface: Slag = slag Δ t =f(stirring energy of metal) ~1/10   Effective Equilibrium Reaction Zone Model(EERZM) 7 Elmira Moosavi- Khoonsari, M.-A. Van Ende and I.-H. Jung: Metall. Mater. Trans. B ., 2022, vol. 53 B, pp. 981-998 If of each species same  

8. Work Plan Commercial Powder injection process Dividing into finite number of Reaction zones (Transitory +Permanent FACTSAGE Calculating equilibrium in each reaction zone EERZ model Accurate Thermodynamic database Kinetics Overall mass transfer coefficient Kinetic Simulation Model Major Predictions S variation in HM versus time Top slag phase and chemistry evolution Influence of Reagent amounts of S variation Sampling from Plant Tuning based on plant data Process Simulation Model 8

9. Modelling of Plume Geometry

10. Plume Volume Calculation L: lance immersion depth(m) D: Diameter of Ladle(m) H P : Plume Height(m) δ : Plume radius C: plume volume correction factor Q: Gas flow rate(Nm 3 /sec)

11. R-1 Reaction-Dissolved Mg and HM in plume Main Assumptions for R-1 reaction: Magnesium vaporizes instantly Homogeneously distributed in bubble plume Rate of dissolution of Magnesium is rate controlling Magnesium efficiency is the quantitative estimation of dissolution rate of Mg Magnesium injected Magnesium lost Dissolved Mg in HM Dissolved Sulphur   Solubility product of MgS =  

12. R-2 Plume and lime particle reaction Hot metal reacting = k m t (n p A p ) ρ m Δ t Lime particle reacting = k p t (n p A p ) ρ p Δ t No. of lime particles n p = 6.W.t/ Π d p 3 ρ p t = H/U m U m = 19.9.(Q/D 2 )(gD 5 /Q 2 ) 0.24 (L/D) 0.20 (H/L) 0.52 k m t = 2.(D m u/ Π d p ) (1/2) D m = 2.8x10 -4 u = m - p ) 2 g 2 d p 3 /225 ρ m μ m ) (1/3) k p t = (1/100). k m t   Quantity Description k m t Overall mass transfer coefficient for metal (m/sec) k m t Overall mass transfer coefficient for lime particle(m/sec) np No.of lime particles in plume A p Area of single lime particle(m2) d p Diameter of lime particle, micron W Lime particle injection rate(kg/sec) t Residence time of lime particles(sec) H Metal bath depth(m) L Lance immersion depth(m) Q Gas flow rate(Nm3/sec) u Slip velocity(m/sec)

13. Transitory reaction zone important parameters Quantity Known parameters Unknown parameters Plume volume Q(Gas flow rate)(Nm 3 /sec) H p (Height of plume)(m) L(Lance immersion depth(m) D(Diameter of vessel)(m) R-1 P MgS (Solubility product) η Mg (Magnesium efficiency) R-2 Flux injection rate(kg/sec) U m (Plume rise velocity) Diameter of lime particle(micron) t p (sec)(residence time of particle Metal bath depth(m) k m t Density of lime particle(kg/m 3 ) k p t

14. 14 Reactions possible in Transitory reaction zone-Continued Reacted HM(with lime particle) Unreacted HM in plume Homogenized metal plume R-3 reaction

15. 15 Permanent contact reaction zones reactions Top slag     Unreacted lime particles Top slag R-4 reaction New solids in top slag Solids in top slag

16. R-5 Plume and top slag reaction Hot metal reacting(kg) = k m p A ρ m Δ t Top slag reacting(kg) = k s p A ρ s Δ t k m p = 2.18X10^(-3)X(L 2 ϵ /D) (1/2) k s p = 1/10(k m p ) Quantity Description k m p Overall- mass transfer coefficient for metal for permanent contact reaction(m/sec) k s p Overall- mass transfer coefficient for slag for permanent contact reaction (m/sec) A Top-slag and HM interfacial area, m 2 L Lance immersion depth, m ϵ Stirring energy(W/kg)

17. R-6 &R-7 Reactions in Permanent zone R-6 Reacted plume Unreacted plume Homogenized plume Homogenisation rection R-7 Reacted slag Unreacted slag Homogenisation rection Homogenized slag To next time step

18. R-8 Permanent contact zone reactions Exchange reaction If t mix < ∆t calc-step Plume reacted = Total Plume volume Remaining HM = Total remaining HM If t mix > ∆t calc-step % Exchange mass-variable 100%Plume 100%Remaining HM If t mix < ∆t calc-step HM new If t mix > ∆t calc-step α % V plume α % V rem_Hm Exchanged Plume Exchanged HM

19. Development of model on a test plant data(two heats) Process parameter Heat 1 Heat 2 Hot metal mass(ton) 288 283 Hot metal temperature( o C ) 1370 1399 Slag mass(ton) 2 2 Slag density(kg/m 3 ) 2800 2800 Metal bath depth(m) 3.7 3.7 Vessel Diameter(m) 4 4 Lime- particle diameter(µm) 110 110 Lime-injection rate(kg/sec) 1.74 1.73 Lime injection period(sec) 420 420 Magnesium injection rate(kg/sec) 0.39 0.48 Magnesium injection period(sec) 300 300 Gas flow rate (Nm 3 /sec) 0.28 0.28

20. Development of model on a test plant data HM composition(wt.%) Element Heat 1 Heat 2 C 4.30 4.30 Si 0.354 0.474 Mn 0.419 0.421 P 0.066 0.066 Cr 0.025 0.025 S 0.0203 0.0229 Fe Bal. Bal. Top Slag composition(wt.%) CaO 38.80 MgO 9.0 Al 2 O 3 14.60 SiO 2 34.60 CaS 2.40

21. Estimation of Reaction volumes at each time step Parameter Heat 1 Heat 2 Plume volume(m 3 ) 17.1 17.1 Mixing energy(W/kg) 0.7 0.7 Mixing time(min) 0.9 0.9 Metal rising velocity(m/sec) 6.04 6.04 Particle residence time(sec) 0.6 0.6 Lime particle slip velocity(m/sec) 83.0E-04 83.0E-04 k m t 5.57E-04 5.41E-04 k m p 3.43E-04 3.49E-04

22. Calculation procedure Preparation of equilib files for various streams Hot metal Top slag+CaO+Mg+N 2 Preparation of Equilib files for all the reaction zones Creation of Macro file to automate the process. Initialization of various streams from data input through excel file. Loop calculation for equilibrium at each reaction zone at each time step Saving the output by suitable thermochemical variables Write the outputs to a excel file  

23. Screenshots of Equilib files R-1 Reaction R-2 Reaction

24. Screenshots of Equilib files R-3 Reaction R-4 Reaction

25. Screenshots of Equilib files R-5 Reaction R-6 Reaction

26. Screenshots of Equilib files R-7 Reaction R-8 Reaction

27. Input data for Heat 1 Transitory reaction zone Permanent contact reaction zone Time Mg(kg) CaO(kg) %Mg reacted %Plume reacted %CaO reacted %plume reacted CaO %Plume reacted %Slag reacted %plume mass exchanged 103.8 19.27 3.55 15.81 37.85 100.00 1 28.8 103.8 43 100 19.27 3.55 15.81 37.85 100.00 2 28.8 103.8 43 100 19.27 3.55 15.81 37.85 100.00 3 28.8 103.8 43 100 19.27 3.55 15.81 37.85 100.00 4 28.8 103.8 43 100 19.27 3.55 15.81 37.85 100.00 5 28.8 103.8 43 100 19.27 3.55 15.81 37.85 100.00 6 28.8 103.8 43 100 19.27 3.55 15.81 37.85 100.00 7 103.8 19.27 3.55 15.81 37.85 100.00 8 15.81 37.85 100.00 9 15.81 37.85 100.00 10 15.81 37.85 100.00

28. Simulation Results Heat 1 Heat 2

29. Validation of model from plant data Heat 1 Model prediction vs plant data

30. Separate contribution to desulphurization Simulation based on Heat 2 conditions

31. Application of model to plant operation Simulation based on heat 1 parameters

32. Optimized process estimation

33. Conclusions The model considers the contribution of both transitory and permanent contact reaction zones to the desulphurization process The developed model was successfully applied to model plant data Developed model also predicts successfully the changes in sulphur wrt to time Sc-1 & Sc-5 seems to be the optimized reagent addition for the process

34. Brief Bio Name – Vaibhav Mishra Educational & Experience background B.Tech –Metallurgy and Materials Engineering, NIFFT, Ranchi( 2014-18) GET & Senior Executive in CI Rings Foundry in Tenneco-Bangalore(2018-2022) M.Tech-Materials Science and Engineering, IIT Kanpur (2022- Present)

35. Literature survey on important modelling studies Year Study Modelling approach Reagent Outcomes 1997 Seshadri et al. Reaction interface modelling CaO-based Effect of process parameters on rate of desulphurization 2016 Visser(PhD Thesis) Reaction interface modelling CaO+Mg(No top slag modelled) Dissolved S vs time( tested with plant data) 2017 Ma et al. Reaction interface modelling (40%)CaO+(60%)Mg(Top slag modelled) Dissolved S vs time( tested with plant data; reagent consumption optimized 2022 Moosavi-Khoonsari et al. Effective equilibrium reaction zone approach CaO+Mg(Top slag modelled) S vs time; phase evolution of top slag Reagent optimized

36. Total Interfacial Area Required Slag Hot metal Interface 12.56 m 2 15294 m 2 Required!!