2017 Volume 57 Issue 10 Pages 1788-1796
Growing concerns over fossil CO2 emissions has created a considerable interest in an efficient utilization of renewable biomass in steel industry. Biomass lignin can be used as binder and reducing agent in the blast furnace briquettes. The traditional briquettes consist of various iron oxide-containing residues and cement is used as binder to give the proper mechanical strength. In the present study, cement (C) has been partially and totally substituted with lignin (L) to produce briquettes containing 0–12 wt.% lignin (L/C: 0, 10, 25, 50 and 100%). The mechanical strength has been evaluated based on drop test and tumbler index measurement. The partial replacement of cement with lignin up to 25% (3.0 wt.% lignin in briquettes) was exhibited adequate briquettes strength for blast furnace application. At higher substitution rate (L/C: 50 and 100%), the briquettes strength was sharply decreased. The briquettes with proper mechanical strength (L/C: 0, 10 and 25%) were subjected to self-reduction under inert atmosphere using thermogravimetric technique (TGA). The reduction rate of briquettes increased when increasing the cement substitution with lignin. The reduction took place in two main steps at 500–800°C and 800–940°C. Combined effect of gas diffusion and interfacial reaction were the rate determining step at the first stage while carbon gasification was controlling the second step of reduction. Interrupted reduction tests have been conducted to evaluate the compression strength after reduction. For all briquettes, the increased reduction temperature and lignin content deteriorated the briquette’s mechanical strength due to the effect of dehydration and lignin gasification.
The conventional process of steelmaking in an integrated steel plant involves series of interconnected units (coking ovens, agglomeration plants, blast furnaces, basic oxygen furnaces, continuous casting, hot and cold rolling) to convert the iron ore to steel products. Each process generates kind of residues which are not suitable for further processing to final product. These secondary materials include flue dust, filter dust, sludge, de-sulphurized scrap, LD dust, coke fines and mill scales. These materials contain valuable amount of iron and carbon but it is not suitable for direct charging into blast furnace ironmaking without pretreatment such as agglomeration. Some of steel plants are recycling these materials during sintering of iron ore while other companies have briquetting plant which can efficiently convert such secondary materials into cold-bonded briquettes with proper mechanical strength for top charging into the blast furnaces.1,2,3) Under normal conditions cement is used as a binder to give the proper mechanical strength of the briquettes and then it goes to slag during the production of hot metal.4,5,6) Coke and coal are the main sources of fuel and reducing agent in the blast furnace, at the same time they are responsible for the main fossil CO2 emission in steel industry. Intensive work is being done to mitigate the CO2 emissions in steel industry by partial substitution of fossil fuels (coal and coke) with renewable source of energy such as biomass.7,8,9,10,11,12,13,14,15) One of biomass materials which could be used as binder for briquettes and reducing agent in blast furnace is lignin. Lignin is considered as the second most abundant renewable biomass on the earth after cellulose and it represents about 30% of the total raw biomass.16) Lignin is insoluble in water and it is stable in nature and acts as glue by connecting cellulose and hemi-cellulose and consequently gives the rigidity for the cell wall. Lignin can be extracted from the process flow in pulp and paper industry by treating the black liquor which contains about 40% of lignin. The pulp and paper mill generates about 1.8 ton of black liquor per ton of pulp product. The annual global production of black liquor from the pulp and paper industry is approximately 215 million tons.17) The black liquor is often used internally in the pulp mill as a source of energy but at low energy efficiency. Recently, an advanced LignoBoost technology has been developed by Innventia AB in Sweden for the separation of lignin from the black liquor.18) The Lignoboost technology can lead to significant positive impacts on exploiting the surplus energy in the pulp mill and enhancing the production capacity. The recent evaluations indicated that LignoBoost technology is able to provide many benefits to the pulp/paper industry such as off-load boiler recovery, replacement of fossil fuels in limekiln and export of lignin biofuel to external users. These benefits are highly value add to the pulp mill industry and it will provide a strong potential for more investiment in lignin. The partial replacement of cement with lignin in the blast furnace briquettes is able to enhance the synergies between steel industry and pulp/paper industry and it could achieve many benefits for both sides. The main objective of the present work aims to evaluate the efficiency of lignin to be used as a binder and reducing agent in the blast furnace. A briquetting machine (Teksam VU600/6) has been used for briquettes production. The mechanical strength has been measured based on drop test and standard tumbler index. The reduction trials of the briquettes have been conducted in thermogravimetric analyzer (TGA) coupled with quadruple mass spectroscopy (QMS). XRD has been used to analyze and investigate the mineralogical composition of briquettes before and after reduction. Compression strength device has been used to measure the briquettes strength after reduction.
The briquettes have been prepared from in-plant fines which are generated in the integrated steel plant. The in-plant fines composed of desulphurization scrap, converter sludge, briquettes fines, filter dust and blast furnace dust. These materials have been provided by SSAB MEROX at Luleå and mixed in specific proportions to prepare a homogenous mixture of in-plant fines. Kraft lignin extracted from pulp and paper industry has been provided by Innventia AB. The chemical composition of in-plant fines mixture, cement and lignin is given in Tables 1, 2. Lump size Kraft lignin provided by Innventia AB is grinded to sizes <1.0 mm and then dried for 72 h at 60°C to improve its mixing with in-plant fines residues. The water content of lignin is reduced from about 37% to 11% by drying. Five recipes with different lignin/cement (L/C) ratios, as given in Table 3, have been designed to investigate the binding and reduction potentials of lignin in the blast furnace briquettes. Specific amount of in-plant fines, cement and lignin are mixed together in laboratory mixing machine according to the proportions illustrated in Table 3. The water is gradually added until the mixture start to stick together to form a paste-like slurry. The mixture is removed from mixing machine after about 10 minutes and then further hand mixing is conducted in order to verify the good homogeneity of the briquette`s mixture. The prepared mixture is then fed into special moulds of laboratory briquetting machine of type TEKSAM VU600/6. The briquetting machine contains a hydraulic station and a press up to 120 bars. The briquettes steel mould is installed on a vibrating station to improve and enhance the briquetting performance. The steel mould is filled with briquette’s mixture and vibration is applied followed by mechanical pressing force to compact the paste material in form of briquettes with hexagonal shape as can be seen in Fig. 1. The weight of the briquettes was in range of 450–500 g. The briquettes are left for drying in the air atmosphere at room temperature. Drop test has been conducted after different curing periods (7, 14, 21, and 28 days) in order to evaluate the mechanical strength of the briquettes. The strength of briquettes was determined by the free fall of the briquette on a steel plate from 1.0 m height and then counting the number of free falls which did not cause sever destruction of the briquette. The drop test is stopped when half of the tested briquette is broken, but the maximum number is 50 drops. Beside the drop test, standard tumbler index (TI, Swedish Standard ISO 3271/2007) has been measured using tumbling machine which is usually applied to determine the disintegration tendencies of blast furnace agglomerates. About 15 kg of the briquettes are charged to a cylindrical drum and left for rotating at speed of 25 rpm for 8 min. The briquettes are then removed and sieved using 6.3 mm screen. The percentage of size +6.3 mm represents the TI%.
Photomacrograph and schematic diagram of the prepared briquettes.
The briquettes with proper mechanical strength have been subjected to a self-reduction in an inert atmosphere (Ar) using TGA. The TGA is Netzsch thermal analysis STA 409 instrument with sensitivity ±1 μg coupled with a quadruple mass spectroscopy (QMS). The sample (weight ~4 g, size ~3.4 cm3) was placed on a Pan ‘like’ crucible and inserted to the hot zone of the furnace. The sample was heated according to the thermal cycle given in Table 4 to simulate the temperature profile in the blast furnace shaft. A continuous flow of Ar was maintained at 200 ml/min (gas velocity ~1.0 m/min) during the whole course of reaction. In order to evaluate the mechanical and mineralogical properties of the briquettes as the reaction progresses, an interrupted test was conducted. Three samples of each type were hanged in a vertical tube furnace. The samples were heated under similar conditions to the TGA; thermal profile and inert gas. The reaction was arrested at predetermined position by removing the sample from the hot zone and quenching in a stream of cool inert gas. The mechanical strength of the briquettes during and after reduction has been measured using cold compression strength device and then investigated using X-ray diffraction analysis (Analytical Empyrean XRD).
The chemical analysis of briquettes from different recipes is given in Table 5. The total iron content was not significantly changed while calcium oxide and silica are decreased from recipe 1 to 5 due to the higher contribution of lignin on account of cement. The carbon and sulphur content of briquettes have been measured using LECO CS230 analyzer as given in Fig. 2. It can be seen that the carbon content increased from about 8.0 wt.% in recipe 1 (0% lignin) to about 15.5 wt.% in recipe 5 (12% lignin). On the other hand the sulfur content was in the range of 0.9–1.08 wt.% for different types of briquettes and the sulphur content in recipe 5 is lower than that of recipe 1. The apparent density of the briquettes has been measured as shown in Fig. 3. The apparent density decreased as the substitution rate of lignin for cement increased. This is attributed to the lower density of the biomass lignin (630–720 kg/m3) compared to that of cement (~1600 kg/m3).
Carbon and sulfur analysis of briquettes from different recipes.
Apparent density measured for briquettes from different recipes.
Drop test has been conducted for the briquettes after different curing periods as can be seen in Fig. 4. Recipes 1–3 (0, 1.2 and 3 wt.% lignin) were able to reach 50 drops without complete broken after 28 days of curing. Recipes 4 and 5 which have been prepared with higher substitution rate of cement with lignin (50C/50L and 0C/100L) exhibited lower mechanical strength and the number of drops was not enhanced even after 4 weeks of drying. A comparison between the weight loss during drop test of recipes 1–3 after drying for 28 days is given in Fig. 5. The briquettes of recipes 1 and 2 (L/C= 0/100 and 10/90) showed similar trend with total weight loss of about 13–15% after 50 consecutive drops. On the other hand the briquettes of recipe 3 (L/C= 25/75) demonstrated higher total weight loss (~36%) compared to that of recipes 1 and 2. The weight loss is sharply increased for the briquettes of recipe 3 after 20 consecutive drops. The tumbler index (TI) has been measured for briquettes of recipes 1–4 using tumbler device as can be seen in Fig. 6. About 15 kg of briquettes from each recipe are tumbled in a rotating drum with a diameter of about 90 cm for 200 revolutions at rotation speed 25 rpm. The briquettes are then screened with sieve of 6.3 mm. The weight percentage of fractions +6.3 mm represents TI. Recipes 1–3 have demonstrated a very good mechanical strength with TI of about 82.5%, 81.5% and 74%; respectively. On the other hand the briquettes of recipe 4 (L/C= 50/50) was very weak with TI of about 18%. Photomacrographs of recipes 1–4 briquettes after tumbling are given in Fig. 7. The briquettes of recipe 1–3 are just lost the sharp edges while almost all briquettes of recipe 4 are disintegrated to fines fractions. This indicates the possible replacement of up to 25% of cement with equal percentage of lignin without impairing the briquettes strength. At higher substitution rate of cement with lignin (50–100%) the briquettes strength becomes very low and could not reach the minimum level of TI (≥ 60%) for blast furnace implementation.
Drop test for briquettes with different cement/lignin ratios.
Weight loss generated during the drop test of briquettes after drying for 28 days.
Tumbler index for briquettes of different recipes after 28 days of drying.
Briquettes after tumbling test of: (a) Recipe 1, (b) Recipe 2, (c) Recipe 3, (d) Recipe 4.
The briquettes from recipes 1–3, which exhibited an adequate mechanical strength for blast furnace, have further subjected to a reduction in Ar using TGA coupled with QMS. The heating profile from room temperature up to 1100°C was selected to simulate the blast furnace shaft as given in Table 4. The weight loss as a function of time and temperature has been recorded as given in Fig. 8. It can be seen that, in all cases ~7% of original weight is lost at temperature lower than 550°C which can be attributed to the evaporation of mechanical and chemical combined water. As the temperature increases the briquettes showed different rate of weight loss. The weight loss rate increased as the percentage of lignin in the samples increased and it follows the order of R3 (3.0% L) > R2 (1.2% L) > R1 (0% L). The total weight loss was increased from 21.7% in R1 briquettes to 26.16 in R3 briquettes. Assuming all lignin (3.0%) in R3 briquettes is consumed during reduction, lignin increased the reduction by about 1.5%. The maximum weight loss difference (6.6%) between R1 and R3 samples can be observed at ~800°C which indicates that lignin has improved the reduction of iron oxides by about 3.5%. The devolatilization of lignin has been conducted separately under similar conditions of temperature and gas atmosphere up to 900°C as shown in Fig. 9. It can be seen fast weight loss during the devolatilization of lignin at temperature lower than 500°C. At temperature 500–800°C, the weight loss becomes very low and then it reached a plateau at temperature 800–900°C. The release of more than 95% of volatile matter content from lignin below 500°C is corresponding to approximately 0.7% and 1.8% weight loss difference in 1.2 and 3% lignin containing briquettes, respectively. Thermogravimetric analysis of Portland cement under similar conditions was showed that up to ~8% of initial weight was lost at 500°C and ~13% weight loss was observed at 900°C. The dissociation of cement in the briquettes would correspond to 1.0%, 0.9% and 0.7% below 500°C for R1, R2 and R3, respectively. Up to 900°C the weight loss was ~1.5%, 1.4% and 1.2% as a result of cement dissociation in R1, R2 and R3, respectively.
Reduction curves of briquettes under inert atmosphere in TGA.
Devolatilization of lignin up to 900°C in an inert atmosphere.
The rate of weight loss of the briquettes is further clarified from the first derivative of the thermogravimetric curve (DTG) as shown in Fig. 10. The TG curves of R2 and R3 (1.2% L and 3% L) showed higher weight loss at relatively low temperatures (550–700°C). The recorded weight loss at ~750°C can be attributed to the reduction of Fe2O3 and Fe3O4 to FeO by either the adjacent carbon, the released volatiles or locally produced CO in the briquettes. With further increase in the temperature (850–900°C), the rate of weight loss increased due to the reduction of FeO to metallic iron and higher gasification rate of carbon. As the embedded carbon consumed and the available iron oxide reduced in the briquettes to metallic iron, the rate of weight loss is gradually decreased.
Derivative thermogravimetric analysis of briquettes in an inert atmosphere.
The off-gases released during the reduction process of R1–R3 briquettes have been analyzed as shown in Figs. 11(a)–11(c). In all the cases, water is evaporated at 100–300°C. Reference briquettes (0% lignin) showed very small peak of H2 at 550°C while lignin samples demonstrated a wide peak of H2 at 20–500°C. This can be attributed to the devolatilization of lignin as the temperature increased as previously described in Fig. 9. It can be seen that the reduction started at around 700°C by generation of CO and CO2. The intensity of CO gas increased with lignin and its two peaks could be related to the stepwise reduction from hematite and magnetite to wüstite (first step) and from wüstite to metallic iron (second step). It can be seen that CO and CO2 are started earlier in the samples containing lignin which can be attributed to the higher reactivity of lignin carbon compared to coke carbon.
Off-gas analysis during the reduction of the briquettes of: (a) Recipe 1, (b) Recipe 2, (c) Recipe 3.
The partially and completely reduced briquettes from R1 and R3 have been subjected to XRD analysis as given in Figs. 12 and 13, respectively. It can be seen that the dried briquettes before reduction are mainly composed of iron oxide and calcium hydroxide phases. As the temperature increased to 750°C, the reduction of Fe2O3 takes place to Fe3O4. At 850°C, the iron oxides are reduced to metallic iron. At 1100°C, the metallic iron became the dominant phase in the reduced samples beside slag (CaO, CaSiO4 and CaMgSiO4). The intensity of metallic iron in lignin samples (R3) is higher than that in reference sample (R1) which confirms the positive impact of lignin on the reduction rate of iron oxides. Moreover, the intensity of slag phases in lignin (R3) sample is lower than that in reference (R1) sample after the reduction at 1100°C. This is attributed to the partial substitution of cement with lignin which reduces the gangue oxides (CaO and SiO2) and consequently the generated slag.
XRD analysis of recipe 1 (reference briquette) after reduction at different temperatures.
XRD analysis of recipe 3 (3 wt.% lignin briquette) after reduction at different temperatures.
Reduction of iron ore to metallic iron is known to be stepwise process with magnetite and wüstite as intermediate products. Reaction of briquettes of self-reducing property proceeds though intermediate gas phase, as the direct contact between solid iron oxide and carbon particles is limited.19) The possibility for solid-solid interaction exists under high vacuum or high flowrate where the locally produced reducing gas is removed immediately.20) The reducing gas (CO) is generated in the district of carbon particles and is transported to the iron oxide particles, resulting in CO2 and lower iron oxides. Subsequently, CO2 gas reacts further with a carbon particle to generate CO. It has been reported that in self-reducing agglomerates, reduction of iron oxide is likely to initiate by solid carbon or volatile matter and propagate by the locally produced CO.21) Therefore, reduction of iron oxide within self-reducing mixture can be summarized in Eqs. (1), (2).
Thus, the overall reaction kinetics in this case depends on gaseous reduction of iron oxide and gasification of carbon. Moreover, some other physical aspects may have a bearing effect on the process. For example, heat transfer, transport of the generated CO to the reaction interface, and diffusion of the product gas to the sample surface and to the gas stream.22) The rate of each step is dependent on the chemical composition, temperature and sample morphology.
Equation (3) is valid for isothermal experiments. For the present study, the reaction kinetic parameters were calculated based on the assumption that the non-isothermal heating cycle consists of an infinite number of discrete isothermal sections; under the present conditions.24) Based on this concept, the thermogravimetric curve (given in Fig. 8) of each case (R1, R2 and R3) was discretized into small segments where the difference in temperature is very small which can be treated as isothermal. The apparent reaction rate (k) of mass loss was then calculated for these isothermal segments taking into account the remaining carbon in the sample for reduction reaction. By adapting Eq. (3) the apparent reaction rate and thus the reaction kinetic parameters can be described in Eq. (4).25)
(i) Residual water removal and partial devolatilization at temperature around 500°C.
(ii) Reduction of higher oxides (hematite and magnetite) to wüstite in a temperature range 500–800°C.
(iii) Reduction of wüstite to metallic iron at temperature above 800°C.
The kinetic parameters of the second and the third steps thus can be calculated based on the above model and assumptions. The relationship between the logarithm of reduction rate and the reciprocal of absolute temperature at 500–800°C and 800–940°C is given in Figs. 14(a) and 14(b), respectively. The estimated values of apparent activation energy are given in Table 6. In all cases the activation energy at first stage (500–800°C) is lower than that at second stage (800–940°C) which can be attributed to the less energy needed to reduce the higher oxides (hematite and magnetite) compared to that of wüstite. Moreover, at the first stage of reduction the activation energy decreases as the lignin content increases which can be attributed to higher reactivity of lignin carbon compared to that of coke carbon. The diffusion of the gas between solid particles in case of lignin briquettes (R2 and R3) is much easier compared to that in briquettes without lignin (R1). At this stage the rate controlling mechanism is a combination of gaseous diffusion and interfacial reaction.26) As the reduction proceeded to the second stage fewer iron oxides become available for reduction and the activation energy becomes higher. The higher activation energy values (260–272 kJ/mol) can be attributed to the deficient of CO2 generated from the reduction process at this stage (Eq. (1)) and the carbon gasification (Eq. (2)) becomes the rate determining mechanism. In lignin briquettes (R2 and R3), lower amount of iron oxides becomes available for reduction compared to the reference case (R1) and consequently the activation energy is slightly increased. The factors affecting the reduction kinetics and mechanisms of iron ore carbon agglomerates have been thoroughly discussed elsewhere.27)
Arrhenius plots for different recipes (R1, R2 and R3) reduced at: (a) 500–800°C, (b) 800–940°C.
The mechanical strength of the briquettes arrested during and after reduction has been measured using cold compression strength (CCS) device to evaluate the ability of reduced briquettes to resist failure under compressive load at room temperature. The average values of compression strength for R1–R3 are given in Figs. 15(a)–15(c), respectively. In all cases, the strength of the briquettes decreased as the temperature increased. The decreasing of CCS becomes significant at temperatures higher than 500°C which can be attributed to the effect of dehydration and consequently the decomposition of calcium silicate hydrate.28,29) With further increasing in temperature, the reduction of iron oxide takes place by self-embedded carbon and consequently the formation of porous structure. The decreasing of cold mechanical strength becomes more significant in lignin briquettes due to the gasification of lignin and the lower slag formation compared to that of reference briquettes which deteriorates the mechanical strength. Future work will be focused on the hot mechanical strength and reduction disintegration index of lignin briquettes surrounded with iron ore pellets to evaluate the briquettes stability and its influence on the pellets reducibility under blast furnace simulated conditions.
Compression strength of reduced briquettes: (a) Recipe 1, (b) Recipe 2, (c) Recipe 3.
Potential for utilization of lignin as a binder and reducing agent on account of cement in the blast furnace briquettes has been investigated. The replacement of cement with biomass lignin up to 25% demonstrated good mechanical strength briquettes for blast furnace implementation. At higher replacement ratios (50% and 100% substitutions) of lignin to cement, the mechanical strength of the briquettes was deteriorated and could not attain the minimum level required by blast furnace. The replacement of cement with lignin has increased the total carbon content while the sulfur content did not significantly change. The reduction potential of briquettes with acceptable strength has been conducted in an inert atmosphere under temperature profile simulated BF conditions. The reduction rate was increased as lignin content increased which was attributed to the faster gasification of biomass carbon at lower temperature. The rate controlling mechanism at first stage of reduction (500–800°C) was combined effect of gas diffusion and interfacial reaction while it converts to carbon gasification at the second stage of reduction (800–940°C). The activation energy was decreased in the first stage by increasing lignin due to the higher reactivity of lignin carbon compared to that of coke carbon. In the second stage, the activation energy increased due to the deficient CO2 generated from reduction of iron oxides. The mechanical strength of all types of briquettes was decreased as the temperature increased higher than 500°C which attributed to the decomposition of calcium silicate hydrate. Lignin briquettes exhibited lower mechanical strength after reduction due to the gasification of lignin and the lower slag formation compared to that of reference briquettes.
The financial support from PRISMA industrial partners (SSAB, MEROX, LKAB) and academic partner Innventia AB is greatly acknowledged. Special thanks to Anita Wedholm form Merox and Mikael Ahlroth from Innventia for materials support and discussion. RENEPRO project, funded by Interreg Nord is acknowledged for support of research work. The partial financial support from CAMM (Centre of Advanced Mining and Metallurgy at Luleå University of Technology) is greatly acknowledged.