Understanding Pulverized Coal Injection in Blast Furnace
Understanding Pulverized Coal Injection in Blast Furnace
Pulverized coal injection (PCI) is a well-established technology for hot metal (HM) production in a blast furnace (BF). It is practiced in most of the BFs and all the new BFs are normally built with PCI capability. The composition and properties of the coal used for injection can influence the operation, stability and productivity of the BF, the quality of the HM, and the composition of the BF gas. The coals being used for the PCI are described in the article under link ‘http://www.ispatguru.com/coal-for-pulverized-coal-injection-in-blast-furnace/’.
The critical aspects of PCI systems include coal preparation, its storage and distribution to ensure uniform feed of coal to each tuyere without fluctuations in the coal delivery rate and its combustion through lance design and oxygen (O2) injection.
Pulverization of coal is carried out in a single or multiple grinding mills (pulverizers) depending on the requirements. Grinding and distribution of the coal to the injection lances constitute a major operating cost. Coal reclaimed from coal storage is screened for the removal of the foreign material and any large lump of coal is crushed. The coal is then fed into the mill where it is pulverized and dried. Coal of the required size is transported out of the mill by the hot gas stream, collected in a bag filter and conveyed to the storage bins. Grinding and transport are carried out under an inert atmosphere to minimize the risk of ignition of the dry coal particles. The resultant particle size distribution of the pulverized coal affects it handleability in pneumatic transport equipment and, at high injection rates, its combustibility.
Pulverizers grind coal to one of the two size fractions namely (i) pulverized coal where around 70 % to 80 % of the coal is under 75 microns (micro metres) and the rest is below 2 mm, and (ii) granular coal which has a 2 mm to 3 mm top size with a limit of 2 % of coal over 2 mm and 20 % to 30 % below 75 microns. Systems injecting this coal size are termed granular coal injection (GCI). The coarser grinding has the advantage of lower grinding and drying costs with the ground coal easier to handle. The finer grinding has a higher burnout in the raceway. PCI is more popular technology than the GCI.
Coal fineness can be varied in the pulverizer by a number of measures which include varying the coal feed rate, the classifier settings, or the air flow rate. Although mills can be tuned to suit a particular coal to produce the required size, this is not practical in case of a coal blend where a large number of coals are being used. In such case some of the coal constituents may not achieve the required fineness.
One of the functions of the pulverizer is to remove as much moisture from the coal as possible. Drying is necessary as moisture causes the problems of free flow both in the pneumatic transport system and in the storage bins. Also, moisture needs to be minimized since additional energy is required for its removal in the BF and injection of moisture increases the reductant rate. In addition, coals with higher moisture consume more power in pulverizers during their grinding and lower the throughput.
Evaporation of the coal surface moisture prevents agglomeration problems within the pulverizer since coals with high moisture and clay contents are mostly liable to sticking. Hence, it is normally necessary to reduce the total moisture content of the coal to around the equilibrium moisture level to reduce handling problems within the mill and storage bins. The moisture content of coal leaving the mill is to be two thirds of the equilibrium moisture level. Equilibrium moisture of the coal varies with its rank, maceral composition, and ash content.
It is important to ensure that the coal is ground to the desired fineness with minimum wear on the pulverizer parts and with least power consumption in order to lower the operating costs. Wear influences the shut downs and maintenance of the pulverizers. Coal properties which influence the wear include the ash content and composition, particle size distribution, moisture, and bulk density. Higher moisture content of the coal accelerates the wear due to the combined effects of wear and corrosion. The abrasive (hard) minerals in coal ash include silica (SiO2) and pyrite (FeS2). Besides wearing of the grinding elements, the abrasive minerals can also erode the pipes and ducts. The most commonly used test for evaluating the abrasion properties of coal is the ‘Abrasion Index’ (AI). Normally coals with a high AI are expected to cause higher wear rates.
Reduction in the power consumption of the mill lowers the operating costs. The power consumption and capacity (throughput) of the mill depends on its design, mill settings, the required fineness, and the properties of the coal. The higher the reduction in the coal size required, the larger is the power consumption needed. Higher level of the coal fineness needs increased mill capacity which may be also necessary when grinding difficult coals.
The main coal property which has the main influence on the mill power consumption and capacity is the hardness, determined by the ‘Hardgrove grinding index’ (HGI). Usually, the higher is the HGI, the easier it is to grind the coal, with consequent lower power consumption and higher throughput. If the design capacity of the pulverizer is limiting the PCI rate, then it is possible to increase injection rates by switching to a softer coal. Increasing the percentage of low volatile matter (VM), high calorific value (CV) soft coal in the high VM, hard coal blend helps in increasing the pulverizer capacity, as well as lowering the blast pressure in the BF and improving coal consumption.
The maceral composition also affects the grinding. In general, higher vitrinite coals tend to have lower grinding energy requirements than lower vitrinite coals since vitrinite is more easily ground than inertinite and liptinite. The effect of rank decreases above a reflectance of around 1.6 where the required breakage energy for vitrinite and inertinite are about the same.
Coals are generally blended to optimize the relative strengths. However, blends do not behave as an average of their components, but can be affected disproportionately by one coal with problem characteristics. Preferential grinding of the softer coal occurs when blends of two coals whose HGI differs by more than 20 are pulverized. Pulverization of blends of ‘hard’ and ‘soft’ coals have shown that the poor characteristics of the constituent coals tend to dominate the blend, with the pulverizer performance more closely resembling that of the harder coal. Preferential grinding of the softer macerals can also occur when milling blends. Coals containing swelling clays can absorb moisture after they leave the pulverizer and cool down. Even when present as a component of a blend, such coals can lead to blockages in the injection systems.
Coal injection system
The injection system pneumatically transports and meters the pulverized coal from the storage bin through the injectant vessel, where it is pressurized up to or above the BF pressure, to the tuyere injection lances. The lances inject the coal in equal amounts through the tuyeres, which are arranged symmetrically around the circumference of the BF. A critical factor in the distribution system design is to ensure uniform feed of coal to each tuyere without fluctuations in the coal delivery route. Any interruption in coal supply can quickly lead to serious problems. The higher is the injection rate, the more serious is the consequences of an unplanned interruption.
At least two injection vessels are needed to provide a continuous flow of coal to the BF. Basically, there are two different arrangements of these vessels namely (i) serial arrangement where the upper vessel periodically replenishes the lower one, which is always kept under pressure, and injects the coal, and (ii) parallel arrangement where the two vessels inject alternately with an overlapping operation to maintain coal injection during the change-over period.
It is important to control the amount of coal injected. Hence, the injectant vessels are continuously weighed and the flow rate of the coal is carefully controlled. Handling problems of pulverized coal in the storage bins which feed the injection vessels and pipelines are due to the amounts of moisture and ultrafine particles, and the presence of clays in the coal. External heaters and/or insulation may be needed to reduce the likelihood of bin blockages due to the condensation of moisture which can occur on the inside of the bin walls. In some places, nitrogen (N2) is blown through aeration pads in the bottom of the intermediate injection tank to ensure free flowing when pulverized coal is transferred to the lower injection tank.
Coal from the injection vessels is generally transported by (i) individual pipes to each tuyere in which case the amount of coal is independently controlled and charged in each pipe, (ii) a common pipeline to a distributor adjacent to the BF in which case, the distributor equally divides the coal into the individual pipes leading to each tuyere. An advantage with this system is that the distance between the coal preparation plant and BF can be longer than with the individual pipe system.
Differences in the routing of the pipes to the tuyeres and the inevitable uneven splitting of the coal at the splitting points can result in an uneven feed to the tuyeres. Imbalances can also cause uneven wear on the pipes and distributor.
Depending on the ratio of coal to the conveying gas, the coal is pneumatically conveyed from the injection vessel to the tuyeres in either dilute phase, or dense phase. The conveying gas loading in case of dilute phase system for coal is typically around 10 kg of coal per kg of carrier gas, and the carrier gas speed is around 15 meter/second (m/s) to 20 m/s. The carrier gas is normally a mixture of N2 and air with compressed air is added to the pipeline below the injection vessel. In case of the dense phase system, the loading is around 40 kg to 80 kg of coal per kg of carrier gas, and the carrier gas speed is around 1 m/s to 5 m/s. The carrier gas is normally N2 or a mixture of N2 and air.
The carrier gas velocity is always to be higher than the minimum conveying velocity in order to prevent blockages. This minimum velocity depends on a number of parameters which include the system pressure and the pipe diameter. These variables interact with each other. The low velocity in dense phase system means low pipeline and component wear, whereas the high conveying velocity of dilute phase system can lead to wear, particularly at pipe bends. The wear rate is determined by the hardness, shape and velocity of the coal particles. Coal properties also influence the wear. Lining the parts of the pipes prone to wear with, for example, a urethane elastomer material provides abrasion resistance, as well as retarding the build-up of fines that can lead to blockage. Coal properties which are related to transfer line blockages are moisture content and clay minerals.
High moisture in coals and blends can create problems. Thus strict moisture limits on the ground coals are applied. The presence of clays, which swell in the presence of water, may cause problems, especially if there is a pressure drop in the conveying system, and/or if there is the presence of ultrafine particles. As the fines content (lesser than 5.8 microns) of the pulverized coal increases, the pressure drop in the conveying system increases. If the pressure drop goes above a certain value, which is related to the design of the plant, then the blockage can occur. Plugging of the pipelines is experienced due to the buildup of the deposits at bends in the pipes which is normally related to the soft nature of the coal (finer particle size distribution). While injecting the coal, ultrafine coal (less than 10 microns) initiates the process by sticking to the elbow wall, and once a rough surface is formed then the larger particles began to adhere. In addition, preferential grinding of the softer coal in a coal blend can lead to a high proportion of ultrafine particles, leading to the blockage.
Blockages can be prevented by improvements in the pipe layout and distribution system and, in some cases, by adjusting the preparation system (such as the coal pulverizer) to produce a coarser particle size. The injection system usually has procedure for detecting and clearing blockage since it is a common phenomenon. Conveying lines include purge ports where blockages are cleared, typically with high pressure air. A simple and practical test is needed to assess the flowability and handleability of pulverized coal and the coal blends. This enables the difficult materials to be identified before they are used.
The injection lance injects coal into the blowpipe which leads up to the tuyere. The particles are immediately heated by the hot blast, ignite, gasify and burn. The design and placement of the lance influences the combustion efficiency of the coal. Earlier the lances used to be straight steel lances which were positioned at or close to the tuyere/blowpipe interface. Designs incorporating the injection of O2 directly into the flow of the coal particles (oxy-coal lances) and/or ways of generating more turbulence at the lance tip have been developed to improve combustion efficiency. These include (i) coaxial lances (where the coal is injected through the inner pipe and O2 through the surrounding annulus), (ii) high dispersive lances, (iii) bevelled lances, (iv) slit lances, (v) eccentric (non-concentric) double lances, and (vi) swirl lances.
Preheating of the coal to increase combustion efficiency is also practiced. Problems which occurred when coal was first introduced, such as lance and tuyere blockages and melting of the lance tip, have largely been overcome. Blockages are mostly due to the coal being heated to a temperature where they become sticky and adhere to the surface of the injection lances and tuyeres. Ash deposition is minimized by utilizing coal with a high ash fusion temperature (AFT). For all practical purposes, the AFT is to be 50 deg C higher than the hot blast temperature. Lances can also plug if coals with a high fluidity cake near the tuyere tip. This can be overcome by avoiding coals with high caking indices, or by increasing the flow rate.
Positioning the injection lance closer to the tuyere reduces the extent of ash impingement in the blowpipe. Utilization of air-cooled coaxial lances helps in preventing the blockage and erosion, and can prolong the life of the tip. The flow rate of the cooling air is to be minimized to decrease its cooling effect on the combustion of coal. However, blockage of lances can still be a frequent occurrence. There are set procedures for detecting and clearing these blockages before they can cause any problems.
Use of different alloys for the injection lances and limiting the hot blast temperature has also influenced the melting of the lance tip. The durability of a lance is an important operational consideration as it burns up over time.
Raceways are vital regions of the BF even though their total volume usually does not exceed 1 % of the inner volume of the BF. They supply the process with heat and reducing agents. Injection of coal inevitably affects raceway conditions which, in turn, have consequences outside the raceway. Unburnt particles leaving the raceway can cause operational problems such as reduced permeability, undesirable gas and temperature distributions, excessive coke erosion, and an increase in char carryover. The quantity of unburnt char increases with increasing injection rates. Thus the combustion and gasification behaviour of the injected coal in the raceway is an important element for the stable operation of the BF. It is very apparent that the BF can consume more injected coal than that combusted within the raceway since the unburnt material is consumed elsewhere in the furnace.
Coal combustion within BF has been widely studied. The studies have been conducted using bench-scale equipment such as thermal gravimetric analysis (TGA), drop tube furnaces (DTFs) and wire mesh reactors (WMRs). These techniques do not fully simulate conditions within the raceway. The residence time of pulverized coal particles in a DTF, as an example, is of the order of seconds whereas it is around milliseconds in a raceway of a BF. Hence, these techniques are typically used to provide a comparative evaluation of the different types of coals.
Another approach which is followed is the use of specially designed facilities to simulate raceway conditions. These include the injection of a hot blast into a packed coke bed, often termed ‘hot model’. These have the ability to simulate combustion conditions for short residence times of milliseconds, as well as different raceway locations. However, the pilot-scale facilities still do not fully simulate raceway conditions in a BF. As an example, they may not work at pressures close to the tuyere/bustle main pressure. Higher pressures in the raceway increase the coal gasification rate.
A number of computer models are available for assessing the behaviour of the coal in the raceway and elsewhere in the BF. The validity of these models have been questioned since the mechanisms they are portraying are complex and not fully understood. Their accuracy is dependent on the assumptions made and the validity of relationships built into the models. Since the behaviour of the coal is strongly influenced by BF design and operating conditions, as well as the coal properties, a computer model is perhaps only applicable for a particular BF, operating conditions and the same types of coals on which it has been developed and tested. These are the limitations of all these techniques.
Combustion of coal between the exit of the injection lance and the rear wall of the raceway (a physical distance of around 0.7 m to 2 m) occurs at high temperatures (1400 deg C to 2200 deg C), elevated pressures (around 3 kg/sq cm to 6 kg/sq cm and short residence times (10 milli-seconds to 40 milli-seconds for pulverized particles). It is under these severe conditions that a high level of coal combustion is required to be achieved.
The combustion process for coal can be divided into the following steps with some of them being overlapping.
- The injected pulverized particles (lesser than 75 microns) are rapidly heated as they enter the O2 enriched hot air blast. The heating rate is determined by the operational conditions but is around 100 deg C per second. The hot blast temperature is typically 1000 deg C to 1200 deg C and the gas velocity is around 180 m/s to 250 m/s.
- Pyrolysis of the particles takes place to produce non-condensable volatiles (gases), condensable volatiles (tars) and a carbonaceous char. It takes around 2 milli-seconds to 20 milli-seconds to complete devolatilization.
- Ignition and combustion of the VM take place to produce principally CO2 (carbon di-oxide) and H2O (water vapour). This takes a few milli-seconds.
- Partial combustion of the residual char takes place by O2. Char combustion contributes the majority of the heat released during combustion. Unlike the combustion of VM, in which the VM diffuse towards the O2 rich atmosphere (resulting in a large reaction area), the O2 for char oxidation must be transported to the relatively small particle surface. As a result, char oxidation is a slower process. As long as the VM is being released, O2 cannot contact the char surface due to the high stoichiometric requirements of the VM.
- Gasification of the residual char by CO2 and H2O to produce CO (carbon mono oxide) and H2 (hydrogen). This is the slowest reaction of all these processes, and mainly takes place outside the raceway.
It is the combustion characteristics of coal rather than coke combustion that govern the gas composition and temperature distribution in the raceway since they are preferentially combusted. Fig 1 illustrates some of the coal combustion steps occurring within the raceway, and how the gas composition varies. Most of the O2 is consumed near the tuyere nose, whilst a CO2-rich atmosphere is produced in the middle, and a CO-rich atmosphere at the end of the raceway.
Fig 1 Pulverized coal combustion and gas composition in raceway
The extent of combustion (combustion efficiency), and hence the amount of unburnt material transported out of the raceway, depends on several parameters which include (i) properties of the coal, such as the VM content, particle size and density, and (ii) operating conditions, for example, blast gas composition and temperature, and lance position and design.
The graph at Fig 2 showing combustion efficiency and pulverized coal injection rate is based on studies made while investigating the maximum rate for pulverized coal from a carbon balance in the BF using a material and heat balance model.
Fig 2 Combustion efficiency and pulverized coal injection rate
Based on the various investigations carried out, the measures to intensify coal combustion in the raceway are summarized below.
- Enriching the blast with O2. However, the non-linear effect of blast O2 on the degree of combustion is to be taken into account. The increase in the combustion rate becomes smaller as The O2 content increases.
- Preliminary mixing of pulverized coal with O2 before introduction into the tuyere cavity.
- Use of coal blends (usually coals with high and low content of VM) and fuel mixtures to maintain both high combustion degree and high coke/coal replacement ratio.
- Coal injection with Fe oxides (fine iron ore, and Fe containing waste etc.), carbonates, and other O2 rich additives.
- Use of chemical and physical phenomena, e.g. catalytic, polarizing and other effects.
- Optimization of coal grinding, depending on operating conditions and coal properties.
Effect of coal rank
The combustion and gasification behaviour of pulverized coal in the raceway is influenced by its properties. The effect of the properties generally on the flame temperature (FT) and combustion efficiency (CE) is described below.
PCI has a cooling effect on the FT. The FT is an important parameter as it affects the slag and metal chemistry, evaporation and recirculation of the alkali elements present, and the flow of metal in the hearth. It is difficult to measure the FT and so it is generally calculated from an energy balance of the raceway zone. The calculated value is known as the ‘raceway adiabatic flame temperature’ (RAFT), or theoretical FT. RAFT calculations can vary from one BF to another depending on the assumptions made, and so values may not be directly comparable. There is an optimum RAFT for every BF depending on factors such as the burden composition and permeability, coke quality, and blowing rate. Injecting coal lowers the RAFT (compared to all-coke operation) as it promotes endothermic reactions. Low VM and high VM coals lower the FT in the ranges of 80 deg C to 120 deg C and 150 deg C to 220 deg C per 100 kg/tHM, respectively. In general, the higher is the H2/C (carbon) ratio in the coal, the more is the cooling effect. The RAFT also decreases with increasing rate of coal injection. Increasing the blast temperature and/or O2 enrichment, and/or decreasing blast moisture can compensate for the cooling effect of coal.
Combustion experiments under conditions simulating the BF environment have indicated that the CE normally increases with increasing coal VM. HV (high volatile) coals are easily gasified, producing a larger quantity of gas, with a lower CV, and a smaller amount of char compared to low volatile (LV) and medium volatile (MV) coals. Thus, gas combustion is more important for the lower rank coals than char combustion. If gas combustion is incomplete, soot can be formed, and this can lead to deterioration in the BF permeability when it leaves the raceway. Soot has a lower reactivity than unburnt char.
The extent of devolatilization is influenced by the coal particle size, with finer sizes leading to more complete devolatilization. As the coal VM content decreases, the ultimate CE is governed by the char reactions since ignition and combustion of the VM is rapid. Char with a higher reactivity has higher CE. It has been often debated that at the high temperatures occurring in the raceway, chemical reactivity becomes less important since combustion rate is limited by the rate of O2 diffusion to the particle, and burnout time depends more on particle size and O2 concentration. Combined with the short residence time, the effect of char reactivity difference between coals may not be very significant in the raceway. There are other opinions which state that in view of the small particle sizes used (higher than 80 % less than 75 microns in PCI) and the highly turbulent conditions which exist in the raceway, the overall rate of char combustion is normally influenced by the intrinsic chemical reactivity of the char. Char reactivity is certainly important outside the raceway. Under the conditions in the upper furnace, char gasification is likely to be controlled by the rate of chemical reaction. Hence, the overall char gasification reaction rate is likely to be influenced by the chemical reactivity of char to CO2.
In general, char reactivity increases with coal VM, that is, HV coals typically produce more reactive chars than LV coals, and hence a better burnout. There are exceptions as the reactivity of char is influenced by a number of factors which include (i) its morphology (surface area and porosity), (ii) its resultant structure, (iii) its composition, and (iv) the operating conditions. The burning rate and reactivity of the char partly depends on the size of the particle and its pore structure. The pore structure controls the supply of reactive gases into the interior of the coal particle and provides a variable internal surface for reaction.
Char fragmentation, which is influenced by its structure, increases the external surface area. A higher proportion of char particles with thin-walled cavities and higher macro-porosity and macro-pore surface areas are produced at high heating rates. In general, these types of chars tend to fragment more than those with thicker walls and lower porosity, and hence have a higher char reaction rate. Fragmentation can be one of the reasons why some operators find that the VM has little effect on the combustibility of coals. Chars formed from higher rank (LV) coals at high temperatures are normally more ordered and hence less reactive. The development of highly anisotropic char cenospheres with increasing temperature also decreases char reactivity. These coals thus benefits from a lower blast temperature in order to improve combustibility.
Changes in a coal’s maceral composition can account for differences in combustion reactivity, particularly among coals of similar rank. The inertinite macerals is usually considered to be ‘inert’ (unreactive). However it is not as simple as this. Not all the inertinite macerals are, in fact, unreactive, and not all the vitrinite ones are reactive. Vitrinite, inertinite, and even liptinite, can contribute to unburnt C in the carbonaceous residue. It has also been noticed that although inertinite-rich coal chars are basically less reactive than the vitrinite-rich coal chars at 500 deg C, this is no longer important at high temperatures (1300 deg C). It is likely that difference in the combustibility of coals is greatly reduced under the very intense combustion conditions in the raceway.
The combustion performance of coals can be improved due to the catalytic effects of the constituent minerals or retarded by excessive mineral concentration. SiO2 and Al2O3 (alumina) can slow down the reaction rate, whilst calcium (Ca), magnesium (Mg), iron (Fe) and alkali types can improve it, with the catalytic effects more pronounced in lower rank coals. However, the improved combustibility of mineral-rich particles has been attributed, not to catalytic effects, but to favourable diffusion of the reacting gas through the minerals and maceral-mineral interfaces. The lack of a clear correlation between char reactivity and the individual inorganic phases can be related to differences in the influence of temperature on coal mineral transformation. Although coals and chars with a high reactivity are usually preferred, too high reactivity can lead to unstable furnace conditions.
Blending can dilute the unfavourable combustion properties of a coal. But the combustion performance of a blend is more complex than that of a single coal. Each of the coal components devolatilizes and combusts at different temperatures and at different times and their burnout can hence vary considerably. Further, interactions between the various coals in the blend can occur which complicates predictions of the combustion behaviour of the blend. Interactions first occur in the pulverizer where there is the potential for large differences in the size distribution of the component coals, especially if there are significant differences in the hardness of each coal. Disproportionation also occurs, influencing the mineral and petrographic composition of the resultant particles, and the subsequent combustion behaviour.
Interactions between the component coals can increase combustibility of the blend. As an example, the combustibility of LV coals can be improved by blending with HV coals. The HV coal releases more VM helping to form a higher gas temperature field, which then heats up the LV coal. This promotes its devolatilization, ignition and combustion. The synergistic effect is more pronounced when the fraction of HV coal is higher, upto a certain percentage. Under simulated BF conditions, a blend containing around 70 % HV coal having 32.5 % VM and 30 % LV coal having 20 % VM has given the highest burnout.
Particle size effects
The combustion performance of coal is influenced by their particle size. For complete conversion, and hence effective utilization of the injected coal, the heating up, devolatilization, pyrolysis and combustion of the particles need to take place in the period between their entry into the hot blast and the raceway boundary. Normally, greater amount of VM is released with reducing coal particle size. This can facilitate gas phase combustion.
Finer particles have higher specific surface areas and thus higher heating rates. The granular coals releases lower amounts of VM than when they are pulverized. Calculated pyrolysis yields indicate that nearly all the VM from the pulverized coals are released whereas it is incomplete in case of the granular coals. The presence of residual VM in the granular coals affects the subsequent CO2 gasification reactivity of the chars. It has also been shown that the extent of devolatilization in the finer particles (45 microns to 75 microns) is more complete than the larger particles (75 microns to 150 microns). The effect is more pronounced for the LV coal (15 % VM) compared to the HV coal (37 % VM). This is since a higher VM release can result in more soot and tar production, produced from secondary reactions of the volatiles. The reactivity of the soot is lower than that of the unburned char. Thus, the lower is the soot formation; the better is the BF stability.
The CE (or burnout) of coal normally increases with decreasing particle size since a higher surface area is available for reaction. Larger particles require a longer time for burnout. The increase is more pronounced as VM content increases in coals. However, the particle size effect is also dependent on O2 stoichiometry, as well as coal rank (and char reactivity). It has been found that larger particles of coal generally have a higher CE (degree of burnout) at O2/C ratios of greater than 2 (fuel lean conditions) under simulated BF conditions. The smaller particles have higher CE under fuel rich conditions (O2/C ratio less than 2).
The effective use of coal needs operational changes to compensate for alterations in the raceway parameters and their effect elsewhere in the BF (such as the thermal state, slag regime and gas dynamics). Measures to intensify the combustion of coal in the tuyere/raceway region, and hence increase the injection rate include (i) increase the amount of O2 in the tuyeres, and (ii) adjustment in the blast temperature and moisture. There are some other measures taken to improve coal combustion, such as preheating the coal and the use of additives. Further, the choice of particle size, and hence the grinding parameters, can also influence the CE.
Oxygen can be added to the tuyere by (i) enrichment of the hot air blast, (ii) injection through the coal lances, and (iii) separate O2 lances. The addition of O2 results into higher availability of O2 for the participation in the combustion of coal in the raceway. Hence, the CE of the coal increases. However, the influence of O2 enrichment on CE is limited. It has been shown through calculation that the CE increases by around 6.7 % for a HV coal (34.5 % VM) and 3.3 % for a LV coal (14 % VM) when O2 enrichment of the hot air blast is raised from 0 % to 6 % by volume. With higher O2 enrichment, CE can actually decrease due to insufficient mixing. Increasing O2 enrichment increases the diffusion of O2, but diminishes the volume of combustion gas which transfers heat to the coal particles. Thus, there is the non-linear effect of blast O2 content on the degree of coal combustion.
Oxygen enrichment of the hot air blast produces both a reduction in bosh gas flow and a rise in FT. The former effect can help counteract the increase in the burden resistance (lower permeability) and the pressure drop associated with high injection rates. The latter effect can help compensate for the cooling effect of the decomposition of the coal VM. The CO and H2 contents also increase with O2 enrichment, resulting in the improved reduction of the iron ore in the central shaft. The CV of the top BF gas normally improves with O2 enrichment.
The lower limit of O2 enrichment is generally determined by the amount needed to maintain the required RAFT, with more O2 needed as the VM content of the coal increases. If the FT becomes too high, then burden descent can become erratic. Too low a FT hampers coal combustion and melting of the ore burden. The upper limit is dependent on maintaining a sufficient top gas temperature. As O2 is increased, the gas mass flow within the BF decreases, which decreases the heat flow to the upper region of the furnace for drying of the burden. The upper limit of the top gas temperature is also governed by the need to protect the top gas equipment. Other limitations to O2 enrichment include its cost and availability.
The position and design of the injection lance influence the CE and ash deposition in the tuyere. However, oxy-coal lance injection (co-annular injection) can produce an insulating effect around the coal particles, resulting in less coal combustion inside the tuyere. This effect carries over into the raceway, and less combustion is the end result. Lowering the O2 lance injection rate in this case improves the CE.
The key measure for combustion at high injection rate is a high blast temperature. O2 enrichment plays a more important role as a means of controlling gas flow in the BF rather than controlling the coal combustion. Normally, a higher hot blast temperature is a cost effective measure than O2 enrichment since it allows a lower consumption of O2. Increased blast temperature also reduces coke consumption, typically 10 kg/tHM for every increase of 40 deg C with PCI, and lead to a small rise in the raceway depth. A higher blast temperature is normally required as the VM of the coal increases. This is due to the lower char reactivity of the low VM coal.
Lowering of blast moisture can help to compensate for the cooling effects of PCI. If the RAFT becomes excessive, then blast moisture can be increased. Raising the hot blast moisture means more H2 in the bosh gas for iron ore reduction. The optimum RAFT in BF operating with higher H2 content can be lower than the BF operating with lower H2. Also, the blast velocity can be adjusted to not only improve coal combustion, but to maintain the needed length of the raceway zone which is critical for obtaining good conditions in the hearth.
As the injection rate increases, the combustibility of coal tends to decrease resulting in unburnt material (such as char, fines, and fly ash) leaving the raceway. Some of these materials, along with coke debris, collect at the back of the raceway, in the bird’s nest, obstructing the rising gas flow and entrained solids in this area. The majority are swept upwards where they can accumulate under the cohesive zone, decreasing permeability and hence the productivity of the BF.
Changes in the permeability of the lower furnace zone can further affect the HM quality and slag viscosity. The unburnt material tends to collect at positions where large changes in the gas flow occur. Eventually it is entrained into the gas flow, passing through the cohesive zone coke slits, and up the shaft, where it can influence burden permeability, and is finally emitted with the top BF gas. Higher coal injection rates also increase the volume of combustion gases, and hence the gas flow, and change the heat load in the lower part of the BF. In addition, more slag is produced.
The deposition of unburnt fine material is a complex phenomenon consisting of several generation mechanisms, reactions, multiphase flow, buildup and re-entrainment. Different gas flow models have been developed to understand and predict the behaviour of fine material within the BF. With suitable burden charging patterns (such as central coke charging) and the use of stronger coke many of the problems relating to gas flow can be overcome.
Operating experience has shown that most of the unburnt material (char) is consumed within the furnace by the three mechanisms which are (i) gasification with CO2 and H2O, (ii) reaction with liquid iron (carburization), and (iii) reaction with slag. It is advantageous if the unburnt char participates in the ore reduction reactions, thus replaces more of the coke and lowers the amount of unburnt solids in the top BF gas. The three char consumption mechanisms are described below.
The gasification reaction of char with CO2 and H2O begins in the raceway, but because the residence time for fine particles is too short for appreciable reaction, gasification mainly occurs in the BF shaft. The reactions of char C with CO2 (the solution loss or Boudouard reaction) and H2O are slower than char combustion. The char obtained from coal competes with that from coke for CO2 and H2O. Char from coal is more reactive than the char from coke and hence is preferentially gasified. Therefore coke degradation by the solution loss reaction decreases with increasing PCI rates. In general, high VM coal char has a higher CO2 reactivity than low VM coal char. However the char reactivity in case of low VM coal can be improved by blending it with the high VM coal. The CO2 reactivity of coal blends is non-additive.
The reactivity of C in the unburnt char to CO2 and H2O is dependent not only on its surface area (particle size) but also on its structure and composition, as well as the operating conditions in the BF. It has been shown that the CO2 gasification reactivity of coal char increases with temperature upto 1500 deg C, especially between 1300 deg C and 1500 deg C. Complete char gasification usually requires a contact time of around 10 seconds at 1500 deg C. Since the residence time for particles at such high temperatures is too short in a BF, hence char gasification mainly occurs at decreasing temperatures in the furnace shaft.
The properties of char change as it moves up the BF, and thus its reactivity to CO2 and H2O. The reacting atmosphere is not uniform. As an example, the concentrations of CO, CO2, H2 and H2O vary at different locations within the BF. Injection of coal increases the bosh gas H2 concentration. Since the chemical reaction rate of H2 reduction is higher than that of CO, the extent of solution loss reaction diminishes as the bosh gas H2 rises. CO2 and H2O are present in the upper part of the BF due to the reduction of iron ore. Under the conditions here, char gasification by CO2 is expected to be controlled by the rate of the chemical reactions. In the lower part of the BF, char gasification is partly diffusion controlled. Hence, the overall reaction rate of char gasification is probably influenced by the chemical reactivity of char to CO2 in this region. Char reactivity towards CO2 is also influenced by its chemical structure, with less ordered structures being more reactive.
The presence of certain minerals in the char ash, such as Fe and alkalis, can catalyze the CO2 gasification reaction, whereas other minerals, such as SiO2 and Al2O3, can slow down the reaction. These catalytic effects become more prominent for low rank coals. Depending on its composition, ash can also retard the C conversion due to blockage of char particles as a result of increased proportion slag formation in the char particle. In the lower part of the BF, condensed alkalis from the recirculating gases (derived from coal, coke and iron ore) can have a catalytic effect. The loss of C by gasification increases the char ash content.
Carburization of the HM begins in the solid phase within the cohesive zone of the BF, and continues during descent of the metal droplets through the active coke, deadman and hearth zones. Unburnt char and fine material leaving the raceway can contact the dripping liquid metal in the bosh and hearth zones. C and other elements, such as Fe, Si (silicon) and S (sulphur), dissolve from the char into the liquid iron and hence influence the composition of the HM. The dissolution of C contributes to the carburization of liquid iron, and controls the level of char consumption by the HM. It becomes critical when the CE is low. If the HM is close to saturation when it reaches deadman and hearth, the unburnt material cannot be consumed, thus reduces the permeability in these regions. The C can come from unburnt coal as well as from coke. Since the dissolution rate of C from coal char is a slower process than that from coke, C from coke is preferentially consumed.
Carbon dissolution from unburnt char into liquid metal is influenced by the furnace operating conditions and the following factors.
- Char particle size – Unburnt char which maintains its pulverized form reacts very little with the liquid metal and the slag since it cannot penetrate into the liquid. If, however, the char particles are agglomerated into larger particles or captured by the larger pieces of coke, then they behave like bosh coke and carburize the liquid metal up to saturation. However, a tuyere probe sample taken at a BF in Australia indicated that ultrafine coal char particles can react with the dripping liquid metal, and that they are more readily dissolved than ultrafine coke particles. Experiments, though, have shown that the dissolution rate of C from coal char, though at larger particle sizes, is a slower process than that from coke.
- Char structure – Normally, the rate of dissolution improves as the C structure becomes more ordered.
- Char mineral matter – In general, SiO2, Al2O3 and MgO (magnesia) slow the C dissolution kinetics, whilst CaF2 (calcium fluoride) and Fe oxides increase the rate. The effect of CaO (calcium oxide) is less clear. The reaction of Ca (calcium) with S in the metal produces a layer of CaS (calcium sulphide) which can inhibit C transfer. The AFT is also one of the controlling mechanisms which limit the C dissolution. The formation of an ash layer on the carbonaceous material reduces the surface area available for dissolution, thus retarding C dissolution rates. Low AFT allows easy removal of the ash, in the form of liquid slag. This results in constant exposure of fresh C surface to the liquid iron, permitting the mass transfer of C to the liquid iron.
- Liquid iron composition – It changes over time. The C dissolution rate is typically decreased as the C content of the liquid iron increases. Higher S content also retards the C dissolution. Combustion of coal and coke releases sulphur oxides (SOx) which can react with the descending metal and slag.
Unburnt char, ash, fines and coke can interact with the dripping slag. The slag composition changes as it moves down the BF, with the Fe oxide concentration being continuously reduced. The reactions at the interface between the solid char and liquid slag play a major role in char consumption since they influence the kinetics of the reduction reactions and the contact area between the slag and char available for reaction.
Factors influencing unburnt char interactions with the slag include the slag composition, char C content, and char ash content and composition, as well as the furnace operating conditions. Char consumption by slags basically occurs through the following mechanisms.
- Reduction of the Fe oxides in slags by C in the char – The wetting characteristic has a significant effect on the dominant reduction mechanism taking place. The wetting characteristic of slags varies with slag composition, temperature, time, and carbonaceous material. Wetting varies as a function of time since the reduction of Fe oxide in the slag by char, and the dissolution of the char ash components into the slag, results in continuous variations in the slag and char compositions. An increase in temperature normally results in improved wettability at the slag/C interface. Reduction rate usually increases with increasing slag FeO (2 % to 10 %) content and with increasing reaction temperature (1300 deg C to 1600 deg C). In general, coal chars are poorly wetted by slag containing more than 10 % FeO at 1400 deg C and 1500 deg C. A faster reaction rate for coke suggests that coke fine is preferentially consumed before coal char.
- Reduction of SiO2 in slag by C of the char – It is a function of temperature. At temperatures less than 1500 deg C, only reduction of FeO takes place. At higher temperatures, both SiO2 and FeO in the slag are reduced, thus resulting in increased consumption of the char. SiO2 is reduced by C, through gaseous SiO, to Si or silicon carbide (SiC). Self-reduction of SiO2 in the char ash by C can also occur, resulting in further consumption of the char. The reduction kinetic of SiO2 is influenced by the wettability of char by the slag. Wetting behaviour is improved with an increase in slag SiO2 content, and with an increase in temperature (1500 deg C to 1700 deg C). Greater amounts of SiO2 and FeO in the char ash facilitate the slag/C interaction, leading to improved consumption of these oxides through reduction reactions.
- Interaction between components in the slag and char – This leads to the assimilation of char ash components such as S. In addition, the reduction of MgO in slag by char C can lead to further consumption. Self-reduction of the oxides in the char ash by C can also contribute to char consumption.
Slag viscosity has also a role to play. The presence of unburnt char in the slag can interfere with tapping by increasing slag viscosity, whereas absorption of char usually increases the fluidity of the bosh slag. Changes in slag mobility can affect the position and shape of the fluid and cohesive zones. A high viscosity slag around the tuyeres leads to serious gas flow problems. Slag viscosity is a complex function of slag composition, temperature and partial pressure of O2. Also unburnt char, coke, and unburnt ash from the coal can interact with the slag. All of these carbonaceous materials contribute oxides to the slag. In general, higher amounts of SiO2 or Al2O3 (acidic components) increase slag viscosity, whereas a higher basicity (higher CaO or MgO) lowers slag viscosity because of de-polymerization of the silicate network. Slag viscosity decreases with increasing FeO (0 % to 20 %) content at a fixed basicity. Basicity is normally determined by the CaO/SiO2 ratio. Since the slag does not completely absorb the char and ash in the bosh region, bosh slag usually has a higher basicity than tapped slag. The addition of fluxes can help in solving slag formation problems.