Analyzing cleaner alternatives of solid and gaseous fuels for iron ore sintering in compacts machines

Analyzing cleaner alternatives of solid and gaseous fuels for iron ore sintering in compacts machines

Journal of Cleaner Production 198 (2018) 654e661 Contents lists available at ScienceDirect Journal of Cleaner Production journal homepage: www.elsev...

1MB Sizes 0 Downloads 52 Views

Journal of Cleaner Production 198 (2018) 654e661

Contents lists available at ScienceDirect

Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro

Analyzing cleaner alternatives of solid and gaseous fuels for iron ore sintering in compacts machines Jose Adilson de Castro a, *, Elizabeth Mendes de Oliveira b, Marcos Flavio de Campos a, Cyro Takano c, Jun-ichiro Yagi d a

UFF/RJ e Federal Fluminense University, Graduate Program in Metallurgical Engineering, Volta Redonda 27255-125, Brazil CEFET/RJ e Federal Center for Technological Education, Department of Metallurgical Engineering, Angra dos Reis 23953-030, Brazil ~o Paulo, Department of Metallurgical Engineering, Sa ~o Paulo, Brazil USP/Poli e University of Sa d Tohoku University, Sendai, Japan b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 18 April 2017 Received in revised form 27 June 2018 Accepted 9 July 2018 Available online 11 July 2018

The steel industry has faced challenges with regard to the raw materials and fuels and hence economic and environmental restrictions. This paper is focused on searching alternatives based on biomass and gaseous fuels suitable for replacing the coke breeze fossil fuel. Nevertheless, testing these technologies are expensive. Therefore, comprehensive mathematical models based on transport phenomena are efficient tools to study and indicate new possibilities for designing operational conditions as well as resizing the machines for minimizing the hazardous emissions. We proposed new concept of operation combining gaseous fuels and biomass for partially replacing the coke breeze in a compact sintering machine. Thus three possible ways for 20% of the amount of coke breeze replacement are analyzed in a combined manner: a) replacement by steelmaking mixing gas, b) replacement by biogas and c) replacement by pellets of biomass. The results indicated that about 20% of the solid fossil fuels could be replaced by waste solid residue of biomass (processed as small pellets). For the analyzed cases, the productivity increase of 5% for the steelmaking mixing gas, 15% for the biomass pellets and 25% for the biogas fuel. For all cases considered it was predicted a decrease of the amount of hazardous compounds emissions. © 2018 Elsevier Ltd. All rights reserved.

Keyzwords: Modeling Emission Cleaner sintering process Solid waste Biomass Solid fuel

1. Introduction The steel industry is considered as a carbon and fossil energy intensive industry. About 80 kg of fossil coal per ton of product are consumed at the operation units for raw materials preparation. At the sintering machine only about 50 kg of fossil fuel per ton of sinter are used. Thus, new technologies are welcome with the focus of neutralizing the carbon emissions. The iron ore sinter process is an important operation unit in the integrated steel plant. This step plays an important role at the steel plant, furnishing suitable raw materials for the blast furnace, and usually is responsible for recycling the inner fine dust produced within the whole steelmaking facilities. The size and capacity of the sinter machines vary widely and are mainly limited by the design and capability of the air suction systems. Larger machines, however, present small ability to

* Corresponding author. E-mail address: [email protected] (J.A. de Castro). https://doi.org/10.1016/j.jclepro.2018.07.082 0959-6526/© 2018 Elsevier Ltd. All rights reserved.

handle low-grade raw materials although high-energy efficiencies are usually obtained. The small and compact machines are increasingly becoming attractive due to their ability to use different source of raw materials and low-grade iron ores (Castro et al., 2013a, b; Naglaa et al., 2015). The traditional sinter plant is composed of raw materials preparation, blowing and suction system, sinter strand, cleaning gas, cooling and sinter product classification. Fig. 1 shows a schematic view of the sinter facilities proposed for a new concept of operation integrated with a steel plant combining a large and small blast furnaces. In Fig. 1 the micropelletizer and the auxiliary systems are drawn. The raw materials are received at the dosage system and depending on the product formulation, the materials are fed into a bed and sent to a mixer and micropelletizer, where additives are adjusted. The micropelletizer system is equipped with a screening system that allows the control of the size of the materials charged in the sinter bed. The sintering bed is composed of a hearth sinter layer and the micropellets are charged at the bed

J.A. de Castro et al. / Journal of Cleaner Production 198 (2018) 654e661

Fig. 1. Flowsheet proposal for new operation system of compact iron ore sintering facilities.

uniformly. The Fig. 1 is an actual tested proposal in a steel industry. Actually, this process furnish sinter for two blast furnaces. A large blast furnace and a mini one. This allows the use of variable size of sinter product. The sinter of size over than 5 mm is transported for the large blast furnace screening system and before charged is screened where the size over 15 mm is sent to the large blast furnace charging system while the size between 10 and 15 mm are used as hearth layer and the excess are furnished for the small (mini) blast furnace of about 280 m3. The remained fine materials (less than 5 mm) are charged as returned material for the raw materials dosage system. A typical materials distribution in the original operation (without using micropelletizer and additional screening system) used to be about 20% as hearth bed layer and 15% as fine return to dosage system. The operation with this new proposal has operated with a new material flow distribution, using the outlet strand production as reference, as follows: 10% for hearth bed layer material (10e15 mm), 15% of materials less than 5 mm returned to the dosage system, 20% as product of size between 10 and 15 mm and the remained 55% as sinter product larger than 15 mm. This new distribution was possible due to a better control and larger size charged to the bed of the sinter machine allowed by the new pelletizer system. Thus the new facilities allowed important gains in both, sinter process and blast furnaces charging control. This is possible due to the flexibility offered by the combination of large and mini blast furnaces demand. It is worthy to mention that for all the operation discussed in this study the drum index attained the large and mini blast furnace requirements. Therefore, the sinter product resistance was not a concern. The product gas and air are sucked and passed through a gas cleaning system with electrostatic precipitators and filters before being discharged to the environment. New challenges concerning the outlet gas recycling have been proposed and this paper considers these possibilities as analysis cases to partially reuse the outlet gas of specific wind boxes. Another possibility is the partial replacement of the solid fuels by steelmaking gas such as blast furnace and coke oven gases. This paper explores analysis cases where this technology is considered. Finally, solid fuels replacement by small pellets of biomass produced from fines of charcoal and wood processing are considered. For all cases, however, the need of process adjustments are required. The main process parameters to be controlled are the temperature profile developed and the bonding phases formed, which are aimed at reaching smooth operation with suitable sinter parameters quality such as reducibility and tumbler index. Also a concern are the main hazardous substances and particulates that are produced depending on the operational and raw materials used (Castro et al., 2013a, b; Xu et al., 2018). In this study, the ignition furnace is modified and enlarged

655

dividing ignition and gas burnout zones in order to allow gas fuel utilization and oxygen injection into five wind boxes length aiming to increase the sinter machine efficiency and decrease the specific emissions. The burner furnace is adapted to have two zones: a) Ignition zone using natural gas and b) Gas burner zone using steelmaking or biogas with oxygen enrichment. This concept is suitable for designing different gas utilization systems. The feeder system is adapted to have height control and allows bed adjustments with uniform distributions. Many efforts have been made to develop new technologies aiming at the decrease of the fossil fuels utilization due to the environmental restrictions and also the decrease of the process carbon intensity (Oyama et al., 2011, Guilherme and Castro, 2012). The process is complex and involves various physical and chemical phenomena such as heat, mass and momentum transfer coupled with chemical reactions (Yamaoka and Kawaguchi, 2005; Castro et al., 2012a, 2012b, 2013a, b, Ahan et al., 2013, Kasai et al., 2005). These phenomena take place simultaneously, increasing considerably the complexity of process analysis. Thus, an effective way of developing new concepts and their quantification is the development of comprehensive mathematical models able for handling simultaneously: (a) the mass transfer using reliable rate equations for the chemical reactions, (b) momentum transfer for complex bed structure and (c) interphase heat transfer considering simultaneously convective, radiation and chemical reactions heat transfer. The proposal of this study is to adapt the actual sintering machine to improve the flexibility of the process and allow simultaneously operation with gas recycling, fuel gas utilization, operation with partial operation of mill scale and biomass together with fossil fuels as coke breeze or anthracite. The purpose of this study is to demonstrate the feasibility of using waste biomass from the charcoal technology processing and wood industry by using a micropellets agglomeration and partial replacement of the coke breeze, which has a strong impact on the environmental performance of the iron sintering. This has only a delicate restriction for the actual operation on the integrated steel industry. The demonstration of the feasibility of the proposed technology leads to important improvement on the steelmaking plant. This technology is considered a potential green house mitigation and/or cleaner production. The main focus of the present study therefore is to analyze and propose new design and operational conditions suitable for use in steelmaking by adapting the compact sinter machine in a combined manner, with the pellets of biomass produced using dried and torrefaction processes and recycling gas. 2. Methodology 2.1. Model formulation The iron ore sinter process takes place at a moving strand where air is sucked through the bed transversely while the strand moves. The phenomena that occur within the bed are complex and involve several chemical reactions. A model able to simulate the inner bed have to consider the macroscopic phenomena of heat, momentum and mass transfer with the rate equations experimentally determined. Thus, in this work a mathematical model based on a set of partial differential equations representing the conservations of momentum, energy and chemical species for gas, solid (raw material mix and solidified liquid) and melting phases is presented and applied for simulating the inner bed features. Similar approaches have been used by several authors, with particularities for the detailed mechanisms adopted and the focus of the model development (Cumming and Thurlby, 1990; Mitterlehner et al., 2004; Castro et al., 2012a, b; Ahan et al., 2013; Guilherme and

656

J.A. de Castro et al. / Journal of Cleaner Production 198 (2018) 654e661

Castro, 2016; Zhang et al., 2018). The domain is considered taking into account the control volume of the moving strand with the gas passing transversely through the bed. The set of differential equations, Eqs. (1)e(4), are solved using the boundary conditions that represents the process of gas suction and solid inflow, as well as, the heat losses to the environment by convections and radiation processes. Additional relations accounting for the interphase momentum and energy transfers are presented in Eqs. (5) and (6) while Eqs. (7) and (8) stands for the effects of the softening and melting properties of the raw materials, which strongly effects the bed permeability, heat and mass transfer (Castro et al., 2012a, b, 2013a, b; Zhang et al., 2018).

      v ri εi ui;j v ri εi ui;k ui;j vu v vP ¼ mi i;j  i  Fjil þ vxk vxk vxj vt vxk

(1)

  Nreacts X vðri xi Þ v ri xi ui;k þ ¼ Mn rm vxk vt m¼1

(2)

  vðri εi Hi Þ v ri εi ui;k Hi v þ ¼ vt vxk vxk

ki vHi Cpi vxk

! þ Eil þ

Nreacts X

DHm rm

m¼1

(3)     Nreacts X vðri εi fn Þ v ri εi ui;k fn v vfn þ Deff þ ¼ Mn rm n vt vxk vxk vxk m¼1 !  #  150mg εs εs ug;j   ¼ 1:75rg þ  3 ug;j  us;j  ds 4s ð1  εs Þ ds 4s    us;j  ug;j  us;j

(4)

"

gs Fj

E

gs

(5)

2 3 !1=2   mg Cp;g 1=3 5 kg 4 6εs rjUj 2 þ 0:39 ¼ ðd 4 Þ ds 4s ðds 4s Þ mg s s kg    Tg  Ts

2.2. Numerical features

(6)    εs ¼ 1  0:403½100ds 0:14 1  MAX 0; MIN     Ts  Tm Sm 1; DTm 100

(7)

εg ¼ 1  εs

(8)

εS ¼

X

εi þ εl þ εlS

 ε þ ε 3  l lS dS ¼ dinitial þ dfinal  dinitial εS

stand for the individual solid and gas volume fractions while ds is the single particle diameter, Sm is the shrinkage melting down factor and Tm is the initial softening and melting temperature. These parameters are obtained using standard softening and melting experiment for the raw materials (Nogueira and Fruehan, 2006, Castro et al., 2013a, b, Guilherme and Castro, 2016). The chemical reactions that take place within the sinter bed involve gas and solid reactions. The water vaporization, partial softening and melt and solidification are assumed as controlled by the heat supply and cooling rates while the combustion, reduction and oxidation are assumed temperature and gas composition dependent. The rate equations for such mechanisms can be found elsewhere (Omori, 1987; Castro et al., 2012a, b, 2013a, b, Nagla et al., 2015; Lu et al., 2013). The effect of raw materials composition are included on the rate equations based on empirical data (Mitterlehner et al., 2004, Kasai et al., 2005 and Castro et al., 2013a, b, Kasama et al., 2006, Guilherme and Castro, 2016). The kinetic and thermal data for the new raw materials used in this study, namely micropellets solid fuel produced from waste fines of biomass and wood, are taken into account by considering the ignition temperature, heat of combustion and apparent activation energy obtained by thermogravimetric and differential scanning calorimetric experiments (Rocha et al., 2016). The NOx and SOx formation mechanisms take into account the oxygen potential and the temperature dependency of the rate equations. The PCDD/F net formation use a two steps mechanism of solid surface adsorption and decomposition and formation during cooling which depends on the local gas composition and temperature. The effect of raw materials composition are included on the rate equations based on empirical data (Mitterlehner et al., 2004, Kasai et al., 2005 and Castro et al., 2013a, b, Kasama et al., 2006). The major chemical reactions taken into account in the model is presented in Table 2.

(9)

(10)

The complete set of variables for the phase velocities components, energy and chemical species of each phase is presented in Table 1. The model considers separately the motion of each phase and their respective chemical and physical properties calculated by using mixing or pondered rules. The thermo-physical properties of each phase thermal conductivity (k), heat capacity (Cp), viscosity (m) and specific mass (r) are composition and temperature dependent. The symbols εg and εs

The numerical solution is obtained by using the finite volume method to integrate the differential equations and obtain a set of discretized algebraic equations. The coefficients are obtained using the power law scheme while the momentum and continuity equations of each phase are solved using the SIMPLE (Semi Implicit Method for Pressure Linked Equations) algorithm to resolve the velocity components and pressure simultaneously in a staggered non-uniform grid (Melaen, 1992). The numerical grid of the strand domain is critical for the accuracy of the simulations. Thus, we carried out a continuously refinement using the averaged error less than 1% for the momentum and energy equations as the stop criterion. The final number of control volumes of 22  165  16 was obtained for the cases analyzed in this study. The solution of the set of coupled algebraic nonlinear equations is obtained interactively using the line-by-line iteration procedure based on the tri-diagonal matrix solver algorithm (Patankar, 1985; Melaen, 1992). The convergence criterion was adopted using 103 for the maximum error of all variables estimated for the control volumes, assuring the complete accuracy of the numerical solutions. The computation time depends on the machine used. This model can be run in a personal computer and in parallel versions of multi core machines. All the calculations presented in this study were run in a core I7 machine with average computation time of 5hr. for a complete run.

J.A. de Castro et al. / Journal of Cleaner Production 198 (2018) 654e661

657

Table 1 Phases and chemical species considered in the model. Equations of the gas phase Gas

Equations of the solid phase Solid

Liquid

Momentum: u1,g, u2,g,u3,g,Pg, εg Energy: hg, Tg Chemical species: N2, O2, CO,CO2, H2O, H2, SiO, SO2, CH4 Momentum: u1,s, u2,s,u3,s,Ps,εs Energy: hs, Ts Chemical species: Coke breeze: C,Volatiles, H2O, Al2O3, SiO2, MnO, MgO, CaO, FeS, P2O5, K2O, Na2O, S2 Iron ore (sinter feed): Fe2O3, Fe3O4, FeO, Fe, H2O, Al2O3, SiO2, MnO, MgO, CaO, FeS, P2O5, K2O, Na2O Return Sinter (bed material): Fe2O3, Fe3O4, FeO, Fe, H2O, Al2O3, SiO2, MnO, MgO, CaO, FeS, P2O5, K2O, Na2O Solidified materials: Fe2O3, Fe3O4, FeO, Fe, H2O, Al2O3, SiO2, MnO, MgO, CaO, FeS, P2O5, K2O, Na2O, Ca2Fe3O5, Al2MgO4 Fluxing agente: (MgOCaO)CO3, CaO, H2O, Al2O3, SiO2, MnO, MgO, TiO2 Sinter cake: Fe2O3, Fe3O4, FeO, Fe, H2O, Al2O3, SiO2, MnO, MgO, CaO, FeS, P2O5, K2O, Na2O, Ca2Fe3O5, Al2MgO4 Chemical Species: Intrabed liquid: Fe2O3, Fe3O4, FeO, Fe, H2O, Al2O3, SiO2, MnO, MgO, CaO, FeS, P2O5, K2O, Na2O, Ca2Fe3O5, Al2MgO4

Table 2 Chemical reactions considered in the sintering model. Solid fuels reactions CðiÞ þ O2 ðgÞ/CO2 ðgÞ(full combustion) CðiÞ þ CO2 ðgÞ/2COðgÞ volatilesðiÞ þ a1 O2 /a2 CO2 ðgÞ þ a3 H2 OðgÞ þ a4 N2 ðgÞ volatilesðiÞ þ a5 CO2 ðgÞ/a6 COðgÞ þ a7 H2 ðgÞ þ a8 N2 ðgÞ(I ¼ coke breeze, anthracite and scale) Note: the stoichiometric coefficients depends on the elemental analysis of the solid fuel Carbonates decomposition CaCO3 /CaO þ CO2 ðgÞ Iron oxides reduction/Re-oxidation Fe2 O3 ðiÞ þ CO=H2 ðgÞ/Fe3 O4 ðiÞ þ CO2 =H2 OðgÞ w 3 Fe O ðiÞ þ CO=H2 ðgÞ/ Few OðiÞ þ CO2 =H2 OðgÞ 4w  3 3 4 4w  3 Few OðiÞ þ CO=H2 ðgÞ/wFeðiÞ þ CO2 =H2 OðgÞ 1 wFeðiÞ þ O2 ðgÞ/Few OðiÞ 2 3 1 w Few OðiÞ þ O2 ðgÞ/ Fe O ðiÞ 4w  3 2 4w  3 3 4 1 2Fe3 O4 ðiÞ þ O2 ðgÞ/3Fe2 O3 ðiÞ 2 Water vaporization/condensation/gas equilibrium H2 OðiÞ/H2 OðgÞCO2 ðgÞ þ H2 ðgÞ4COðgÞ þ H2 OðgÞFeSðsÞ þ =O2 ðgÞ4FeOðsÞ þ SOðgÞ þ COðgÞN2 ðgÞ þ 2O2 /2NO2 ðgÞN2 ðgÞ þ O2 ðgÞ42NOðgÞ

1 CðiÞ þ O2 ðgÞ/CO2 ðgÞ(partial combustion) 2 CðiÞ þ H2 OðgÞ/COðgÞ þ H2 ðgÞ

MgCO3 /MgO þ CO2 ðgÞ (i ¼ iron ore, return sinter and scale) (i ¼ iron ore, return sinter and scale) (i ¼ iron ore, return sinter and scale) (i ¼ iron ore, return sinter and scale) (i ¼ iron ore, return sinter and scale) (i ¼ iron ore, return sinter and scale) (i ¼ iron ore, return sinter and scale) SðsÞ þ O2 /SO2 ðgÞSðsÞ þ 1 CO þ O2 /CO2 1 2 O /SOðgÞ 2 2 1 H2 þ O2 /H2 O 2

PCDD/F formation 24 CðsÞ þ 1:5O2 ðgÞ þ 2H2 ðgÞ þ 8HClðgÞ4C12 O2 H5 Cl3 ðgÞ þ C6 H3 C12 OHðgÞ þ C6 H3 Cl3 ðgÞ24 CðsÞ þ O2 ðgÞ þ 2H2 ðgÞ þ 8HClðgÞ4C12 OH5 Cl3 ðgÞ þ C6 H3 C12 OHðgÞ þ C6 H3 Cl3 ðgÞ

3. Results and discussions 3.1. Model validation and verification The model was previously verified for the compact sinter machine using industrial data obtained by inserting thermocouples inside the sinter bed travelling through the entire length. Fig. 2 shows a comparison of the model prediction and the averaged values measured at the outlet of the strand grate. Another temperature record was obtained using the data of the wind boxes. The results and comparison of the averaged wind boxes values are shown in Fig. 3. The raw materials used are listed in Table 3 with their respective composition. The source of sulfur are from the raw materials with chlorine and nitrogen come from the environment.

Both averaged results were compared with the calculated results, showing good agreement, as can be seen in Figs. 2 and 3, respectively. The calculated and measured values were carried out for the compact machine within the bed for the conventional fuel operation (coke breeze), as shown in Fig. 2. The outlet temperatures of the wind boxes were recorded during the entire residence time of the materials within the strand for the new operational and raw materials used. The averaged values were compared with the calculated steady stated conditions of the machine. As can be seen, the model presented quite good agreement with the averaged gas temperature outlet comprising each wind boxes, shown in Fig. 3. Thus, for the analysis cases the model was assumed accurate enough for comparing the different operational conditions. Therefore, three analysis cases were considered, with the goal of

658

J.A. de Castro et al. / Journal of Cleaner Production 198 (2018) 654e661

Fig. 2. Model predictions and measurements for the inner bed temperature on the compact sinter machine for conventional operation.

Fig. 4. Temperature pattern for the case of mixing 20% of coke oven gas and 80% of blast furnace gas replacing solid fuel (coke breeze) e Replacement ratio of 1.5 kg of mixing gas for 1 kg of coke breeze removed from the cake mixture.

and fuel gas used. Therefore the aim of the replacement analysis was to see if the fuel rate was compatible with maintaining the heat requirement for the sintering temperature. 3.2. Cases analysis

Fig. 3. General model verification confronting the wind boxes average outlet temperatures with model predictions for the analysis cases.

3.2.1. Replacement of solid fuel by coke oven and blast furnace gas Fig. 4 shows the temperature distribution pattern inside the sinter bed for the case of fuel gas injection of the sintering machine and using preheating zone with the process gas. The temperature distribution at the combustion zone shows stable conditions and adequate maximum temperatures compatible with the predominant mechanism of liquid formation and solidification, which is characteristic of good quality sintering formation for the iron ore

Table 3 Raw materials used in the analysis cases. Mass % Iron ore Limestone Mill scale Fine dust

C

0.0 0.0 30.5 19.4 C Coke breeze 88 Biomass (pellets) 75 Fuel gas compositions (% vol) Coke oven gas Gas mix Biogas Recycling gas Softening and melting parameters Tm- Starting melting Temperature ( C) Sm- Volumetric shrinkage (%) DTm- Melting interval ( C)

Fe2O3

Fe3O4

Fet

SiO2

CaO

Al2O3

(Ca,Mg)CO3

96.1 0.0 1.2 48.5 volatile 1.5 22

0.8 0.0 2.5 5.3 Ash 11.5 3 O2 0.0 8.8 0.5 16.5 Base case 1248 65 145

67.4 0.0 77.5 56 S 0.0002 0.0006 CH4 14.5 4.8 4.5 0.2 Mixing gas 1232 58 138

1.7 4.5 1.5 2.8 N 0.00005 0.00008 CO 11.5 19.4 33 2.5

0.0 0.0 0.5 4.5

1.3 3.6 1.4 1.5

0.0 91.5 0.5 2.2

H2 53 36.5 48.5 0.5 Biogas 1228 63 142

CO2 8.5 4.5 8 8.3

N2 12.5 26 5.5 71 Biomass 1242 56 137

demonstrating the feasibility of these technologies. The first case considers the possibility of partially replacing the coke breeze by a mixture of coke oven and blast furnace gas, the second one considers the replacement by biogas produced from gasification of biomass and the third the replacement by solid fuel composed of micropellets produced from fines of charcoal and wood processing powder. All these alternative fuels have lower calorific values and faster reaction rates. Thus a replacement ratio was calculated considering the heat released for each fuel compared with replacing some of the coke breeze. Therefore, the aim of the use the fuel rate compatible with the heat requirement for the sintering temperature. The replacement ratio varies depending on the biomass

sinter product, similar with those obtained using coke breeze fuel. This concept is suitable when the steel plant has excess of gas in the coke oven facilities. As can be observed in Table 4, the specific emissions of SOx, NOx and PCDD/F are drastically decreased when this practice was considered. The reason for such behavior is due to the raw materials and changes of the strand inner conditions, which inhibits the formation of these hazardous compounds. 3.2.2. Replacement of solid fuel by biogas (gasification of biomass) In this section, the biogas utilization for replacement of the coke breeze within the sinter bed is considered. Fig. 5 shows the temperature distribution for the condition of coke breeze replacement

J.A. de Castro et al. / Journal of Cleaner Production 198 (2018) 654e661

659

Table 4 Summary of specific emissions for the cases analyzed. Specific Carbon intensity (kg ton1) SOx (SO þ SO2)(ppm) NOx (NO þ NO2)(ppm) PCDD (ng Nm3) PCDF (ng Nm3) Particulates (mg Nm3) Base case Coke breeze Analysis cases Mixing gas Biogas Biomass (micropellets)

68.45

35.32

23.34

0.44

0.91

15.45

73.11 80.75 51.45

1.25 1.24 0.17

20.45 22.13 18.23

0.34 0.01 0.32

0.65 0.86 0.88

16.95 11.28 8.57

ng - nanogram; mg - milligram; ppm - part per million.

velocity of the strand to reach the burnout temperature point of 900  C. As observed in Table 3, the biogas presented a high concentration of H2 and CO, which confers higher calorific value with high combustibility. This features allows intense release of heat within the combustion front and can be used to improve the strand velocity and hence the productivity.

Fig. 5. Temperature pattern for the case of biogas produced from biomass gasification replacing solid fuel (coke breeze) e Replacement ratio 1.6 kg of biogas for 1 kg of coke breeze removed from the cake mixture.

by biogas with adjustment of the oxygen in the ignition and burning furnace. The model prediction indicated that larger high temperature zone is obtained by combining equivalent heat of replacement ratio with faster reactions within the combustion front supported by the additional oxygen. Thus, it is clearly shown that this condition is favorable for the sintering liquid mechanism, which is expected to produce higher mechanical resistance and less return fines. The temperature profiles shown in Figs. 4 and 5 are obtained by interactively adjusting the strand velocity to attain the burnout temperature point. During the calculations, the bed height were constant. Thus the temperature distribution will reflect the reactivity of the fuel and the positions were the injection is carried out. In the case of Fig. 5 compared with Fig. 4, the mix gas has lower calorific value and the oxygen injection plays the major role for distributing the heat at the sintering front. The Biogas (Fig. 5) has large amount of H2 and CO compared with the mixing gas (Fig. 4). Although the total amount of fuel is increased for this case, there are some clear advantages. The reactivity of the fuel gas is higher and the heat supply is well distributed, which allows the sinter strand to increase the velocity keeping similar temperature distribution profile. Another advantage that is highly desirable is that all the replacement and additional fuel is renewable. Thus, this option combines the replacement of fossil fuel by biogas and promotes the improvement of the machine productivity. The best match for the process feasibility and economical return. Table 4 shows the main emissions parameters predicted for this conditions. It is observed that for the biogas utilization the carbon intensity is higher. However, we must point out that this carbon is from biomass, which is recovered by photosynthesis. The specific emissions of the other hazardous compounds is comparable with the mixing gas replacement practice. The replacement ratio for the biogas was 1.6 kg of biogas for each kg of coke breeze removed from the bed mix. In this case the productivity increase by the combination of a slight increase of the apparent bed density and the increase of the

3.2.3. Replacement of solid fuel by micropellets of biomass waste The case of partial replacement of coke breeze by micropellets produced from waste biomass is presented in Fig. 6. In this case 20% of the coke breeze of the cake mixture is replaced by biomass from drying and torrefaction process. The replacement ratio was used with 2 kg of biomass pellets for each kg of coke breeze removed from the raw material mixture. The strand velocity was adjusted in order to keep the burnout temperature point of 900  C. This procedure allowed the productivity increase of about 15%, besides the small decrease on the apparent density of the bed materials. The temperature pattern presented uniform thickness of the sintering zone, which indicates that uniform quality of the final product is expected and hence less return fines materials. Therefore, this operational technique is demonstrated as feasible and preferred. The results presented in this study demonstrated that the evaluated cases are feasible and can partially replace the coke breeze. In addition, the productivity can be increased to account for the faster combustion of the gas and biomasses used. The model predictions indicated that 5% up to 25% of productivity increase could be obtained by varying the biomass to coke breeze replacement up to 20% and corresponding adjustment of the velocity strand, oxygen and replacement ratio. Therefore, these new technologies proposed for the compact sinter machine can significantly contribute for cleaner and sustainable steel industry. Table 4 summarizes the specific emissions and carbon intensity for comparing the environmental impact of the proposed cases. As can be observed, the specific carbon intensity of the process can be substantially decreased with the use of micropellets of biomass

Fig. 6. Temperature pattern for the case of micropellets of waste biomass replacing 20% of the solid fuel (coke breeze) e Replacement ratio 2 kg of biomass micropellets for 1 kg of coke breeze removed from the cake mixture.

660

J.A. de Castro et al. / Journal of Cleaner Production 198 (2018) 654e661

Table 5 Comparison of sinter compositions and measured quality tumbler index.

Base (measured) Base (calculated) Mixing gas (measured) Mixing gas (calculated) Biogas (measured) Biogas (calculated) Biomass (micropellets) (measured) Biomass (micropellets) (calculated)

Fe2O3

Fe3O4

Al2O3

SiO2

CaO

TI-Tumbler index (% > 10 mm)

68.94 69.32 69.15 68.98 68.60 69.40 67.93

10.45 9.97 10.32 9.96 10.48 10.32 10.85

1.71 1.73 1.75 1.72 1.73 1.74 1.74

5.96 5.94 5.80 5.77 5.90 5.84 5.80

9.71 9.74 9.73 9.25 10.20 9.85 10.40

71 e 74 e 73 e 69

68.20

10.82

1.78

5.45

10.20

e

produced with fines of charcoal. The SOx emissions are drastically decreased when compared with the base case of only coke breeze due to low sulfur of the biomass and biogas, as well as the particulates when using biomass pellets. These results are explained due to the complete combustion of the fuel and the low level of ultra fines presented in the sintering mixture with the intensive use of the micropellets system and use of the fine sinter as nucleation materials. Using the coke breeze practice as the reference operation, the carbon intensity parameter decreased about 25% for the best case of 20% of coke breeze replacement by biomass pellets. The replacement of coke breeze by steelmaking mixing gas indicated about 7% increase of carbon intensity. The biogas fuel replacing 20% of coke breeze indicated that 18% increase of the carbon intensity is expected. However, we must emphasize that this carbon is from renewable source, which is advantageous from the point of view of clean technology. From the mass balance analysis, the SOx, NOx and PCDD/F emissions, the proposed analysis cases are clearly advantageous since all analysis indicated that significant amount of these compounds are decreased. Similar trend is observed for the particulates emissions. These patterns are justified due to better control of the in bed gas flowing conditions, since biomass are charged as pellets and the ultra-fine particles are minimized by the micropelletizer system. The sinter produced during the trials runs were measured in order to confirm the composition and strength parameters. The measured and predicted values of the sinter composition for the base and analysis cases are presented in Table 5. A common sinter quality index is measured on the standard tumbler test, which accounts for the remaining amount of þ10 mm after the standard revolution number. The base case represents the actual values of the sinter commonly produced and used in a middle size blast furnace. As can be observed, the values of the TI for the sinter produced under the new operations with the alternative raw materials are acceptable and therefore suitable for using replacing the usual production.

4. Conclusions Alternatives for the fossil fuel coke breeze actually used at the sinter plant of integrated steel works are considered in this paper. The sinter machine is compact and can handle several fuels such as micropellets, gases injection and oxygen enrichment. The analysis are carried out by using a comprehensive mathematical model able to simulate a fuel flex compact sintering machine. Then, the model is used to investigate feasible operations based on steelworks gas, biogas and micropellets made by fine solid waste fuels. Simulation results indicated that feasible operation conditions could be obtained using partial replacement of the coke breeze by steelmaking gas and biogas in a shorter machine with faster strand velocity due to higher combustibility of the fuels. Oxygen enrichment was used

to increase the front combustion and keep the flame stability assuring fuel flexibility. The model predictions indicated that the productivity is increased 5%, 15 and 25% due to the increase of the sinter strand velocity compatible with the temperature pattern of suitable combustion front, which combine the effects of faster reactions with intensive heat supply at the sintering zone. These results were obtained after continuous increase of the strand velocity to keep the same burnout temperature operation point. Acknowledgements The authors thank to the agencies CNPq, CAPES and Faperj for the partial financial support of this research project. The authors thank Dale M. Crouse, Ph.D, for the reading of the manuscript and suggestions to improve it. References Ahan, H., Choi, S., Cho, B., 2013. Process simulation of iron ore sintering bed with flue gas recirculation. Part 2eparametric variation of gas conditions. Ironmak. Steelmak. 40, 128e137. https://doi.org/10.1179/1743281212Y.0000000072. Castro, J.A., Sasaki, Y., Yagi, J., 2012a. Three dimensional mathematical model of the iron ore sintering process based on multiphase theory. Mater. Res. 15, 848e858. https://doi.org/10.1590/S1516-14392012005000107. Castro, J.A., Guilherme, V.S., França, A.B., Sasaki, Y., 2012b. Iron ore sintering process based on alternative gaseous fuels from steelworks. Adv. Mater. Res. 535, 554e560. https://doi.org/10.4028/www.scientific.net/AMR.535-537.554. Castro, J.A., Pereira, J.L., Guilherme, V.S., Rocha, E.P., França, A.B., 2013a. Model predictions of PCDD and PCDF emissions on the iron ore sintering process based on alternative gaseous fuels. J. Mater. Res. Technol. 2, 323e331. https://doi.org/ 10.1016/j.jmrt.2013.06.002. rico da Castro, J.A., França, A.B., Guilherme, V.S., Sasaki, Y., 2013b. Estudo nume ^ncia de propriedades de amolecimento e fusa ~o na cine tica de formaç~ influe ao rio de ferro. Tecnologia em Metal(CaFe2O4-Ca2Fe2O5) na sinterizaç~ ao de mine ~o. https://doi.org/10.4322/tmm.2013.003. https:// urgia, Materiais e Mineraça doi.org/10:16e27. Cumming, M.J., Thurlby, J.A., 1990. Developments in modeling and simulation of iron ore sintering. Ironmak. Steelmak. 17, 245e254. Guilherme, V.S., Castro, J.A., 2012. Utilizaç~ ao de g as de coqueria na sinterizaç~ ao de rio de ferro. Rem Rev. Esc. Minas 65, 357e362. https://doi.org/10.1590/ mine S0370-44672012000300012. Guilherme, V.S., Castro, J.A., 2016. Displacement of the ingnition furnace in the iron ore sintering with Re-Circulation of waste gases. Mater. Sci. Forum 869, 643e648. Kasai, E., Komarov, S., Nushiro, K., Nakano, M., 2005. Design of bed structure aiming the control of void structure formed in the sinter cake. ISIJ Int. 45, 538e543. https://doi.org/10.2355/isijinternational.45.538. Kasama, S., Yamamura, Y., Watanabe, K., 2006. Investigation on the dioxin emission from a commercial sintering plant. ISIJ Int. 46, 1014e1019. https://doi.org/10. 2355/isijinternational.46.1014. Lu, L., Adam, M., Kilburn, M., Hapugoda, S., Somerville, M., Jahanshahi, S., Mathieson, J.G., 2013. Substitution of charcoal for coke breeze in iron ore sintering. ISIJ Int. 53, 1607e1616. https://doi.org/10.2355/isijinternational.53.1607. Melaen, M.C., 1992. Calculation of fluid flows with staggered and nonstaggered curvilinear nonorthogonal grids-the theory. Numer. Heat Transf. B 21, 1e19. https://doi.org/10.1080/10407799208944919. Mitterlehner, J., Loeffler, G., Winter, F., Hofbauer, H., Smid, H., Zwittag, E., et al., 2004. Modeling and simulation of heat front propagation in the iron ore sintering process. ISIJ Int. 44, 11e20. https://doi.org/10.2355/isijinternational.44.11.

J.A. de Castro et al. / Journal of Cleaner Production 198 (2018) 654e661 El-Hussiny, Naglaa A., Khalifa, Ahmed A., El-Midany, Ayman A., Ahmed, Ahmed A., Shalabi, Mohamed E.H., 2015. Effect of replacement coke breeze by charcoal on technical operation of iron ore sintering. Int. J. Sci. Eng. Res. 6, 681e686. Nogueira, P.F., Fruehan, R.J., 2006. Blast furnace burden softening and melting phenomena. Part III: melt onset and initial microstructural transformations in pellets. Metall. Mater. Trans. B 37, 551e558. https://doi.org/10.1007/s11663006-0038-3. Omori, Y., 1987. The Blast Furnace Phenomena and Modeling. Elsevier Applied Science, London. Oyama, N., Iwami, Y., Yamamoto, T., Machida, S., Yguchi, T., Sato, H., Takeda, K., Watanabe, Y., Shimizu, M., 2011. Development of secondary-fuel injection technology for energy reduction in the iron ore sintering process. ISIJ Int. 51, 913e921. https://doi.org/10.2355/isijinternational.51.913. Patankar, S.V., 1985. Numerical Heat Transfer and Fluid Flow. Hemisphere Publishing Company, Washington, p. 197. Rocha, E.P., Castro, J.A., Vitoretti, F.P., Vermilli Jr., F., 2016. Kinetic of self-reducing mixtures of iron ore and biomass of elephant grass. Mater. Sci. Forum 869, 1007e1012. Xu, S., Chen, T., Buekens, A., Li, X., 2018. De novo formation of PCDD/F during sintering: effect of temperature. Granule Size Oxygen Content 1, 1e7 (Online). https://www.jstage.jst.go.jp/article/isijinternational/advpub/0/advpub_ISIJINT2017-392/_pdf. Yamaoka, H., Kawaguchi, T., 2005. Development of a 3-D sinter process mathematical simulation model. ISIJ Int. 45, 522e531. https://doi.org/10.2355/ isijinternational.45.522. Zhang, B., Zhou, J., Li, M., Li, Y., 2018. Modeling and simulation of iron ore sintering process with consideration of granule growth. ISIJ Int. 58, 17e24. https://doi. org/10.2355/isijinternational.ISIJINT-2017-342.

Nomenclature

Variables and symbols

661

A :: surface area, (m2 m3) Cp :: heat capacity, (J kg1 K1) dm :: solid component diameter, (m) ds :: solid phase mean diameter, (m) dinitial :: Initial micro pellets charged, (m) dfinal :: solid agglomerated, (m) Fj :: interaction force in j direction between i and l phases, (Nm3 s1) h :: enthalpy of the phase (kJ kg1) ! U i :: phase velocity vector (i ¼ gas and solid), (m s1) P :: phase pressure (Pa) C $m Prg :: Prandtl number, (), Prg ¼ p;gkg g ! ! r j U g U Sj Regs :: particle Reynolds number, () Regs ¼ g m dS g 1 1 R :: gas constant, (J mol K ) rm :: rates of chemical or phase transformations, (kmol m3 s1) Sf :: source or sink terms for the f variables, (various) Sm :: volume shrinkage in the sintering zone (%) xi :: spatial coordinates, (m) t :: time, (s) T :: temperature, (K) Tim :: initial melting temperature (K) Greek symbols

DTm :: melting temperature interval of the iron ore (K) fn :: mass fraction in Equation (4), (calculated by the model), [kg.kg1] 4m :: solid diameter shape factor (m ¼ sinter feed, sinter return, limestone, fines, coke, mushy and bonding phases), () εi :: volume fractions (m3 m3) ri :: phase density (i ¼ gas and solid), (kg m3) m :: phase effective viscosity (Pa.s)