News

Preparation and Charging of Blast Furnace Burden

• satyendra
• April 26, 2020
• 1 Comment
• automation, bell less top, blast furnace, Burden materials, chute, Conveyor charging, Granular flow, Mixing, Percolation, segregation, stock house,

Preparation and Charging of Blast Furnace Burden

Blast furnace (BF), except at its hearth, is basically a passage for gases and burden particles which move in the counter-current directions in the BF. The basic requirement for stable operation of the BF is to maintain in the furnace a moving layer of burden which does not fluctuate much. Specifically, it is to form a stable gas flow and a burden layer structure free of mixed burden layer. These are closely related to each other. The stability of the gas flow depends almost entirely on the burden permeability, which is determined by the burden packing structure (particle size, particle size distribution, and fine particles ratio etc.), and the burden descent behaviour, which is the solid flow.

In principle, BF processing is a complex counter-current, co-current and/ or cross-current, 4-phase flow system consisting of solids, gases, liquids and powders. The interactions of the phases are distinct and localized in different regions of the BF. There are normally five distinct zones (Fig 1) ina BF namely (i) lumpy zone, (ii) cohesive zone, (iii) coke active zone, (iv) combustion zone (raceway), and (v) dead-man zone (hearth centre region). Due to the differences in burden phases and distinctive interactions in particular regions of the BF, there is no single burden movement or flow pattern in the entire volume of the BF.

Fig 1 Zones of a blast furnace

There are generally four distinct types of solid flow regions in the BF. These are (i) plug flow region which is associated with the uniform velocity of burden descend from stock-line and position just above the cohesive or melting zone, (ii) stagnant flow region (dead-man) which is a discontinuous mass of partially reacted coke particles in the centre part of the hearth, (iii) partly-stagnant flow region which is adjacent to the dead-man where a sluggish movement of partially reacted coke particles is there, and (iv) converging flow region which is around the raceway with significant particles velocity variation within stagnant zone-furnace wall distance.

The mechanism for the burden descend involves the disappearance of ore and coke through their reaction, melting and combustion, the motion of burden particles at the top of the burden layer and near the furnace wall, and the infiltration of fine-grained raw material into a coarse-grained layer etc. The factors which influence the burden descend in the BF include the condition of the raw materials (particle size, strength, burden distribution – ore/coke ratio), the raceway condition (auxiliary fuel injection), and the furnace inner wall profile etc.

BF operation is quite sensitive to the profiling, particle size distribution, and all those factors which affect performance and productivity. For ensuring the smooth operation of BF, it is essential to optimize the distribution of the solid burden materials (iron ore, coke and fluxes) at the stock-line. It is important to control the internal state of the BF through appropriate control of the burden distribution at the stock-line level.

The burden material of the BF is to fulfill certain requirements needed for smooth BF operation. It is to fulfill the mechanical, chemical and thermal requirements. The charged material is to form a strong and permeable structure to carry the burden and allow the passage of the reducing gas in the BF shaft. The early disintegration of the burden is to be avoided since the formation of fines disturb the reducing gas flow in the shaft thereby decrease the process efficiency. Hence, the charging of fine burden materials in the BF is to be eliminated.

The downward movement of the burden, chemical reactions, thermal profiles, and liquid movement in the zones is required to be optimized with the objective of maximizing productivity and ensuring stable furnace operation. Achievement of stable and high BF productivity operation can be realized through optimization of both in-furnace processes and peripheral operations such as thorough control of raw materials physical and chemical properties and particles distribution. Both the in-furnace processes (charged material physical changes and chemical reactions) and peripheral auxiliary operations (burden stockpiling, reclamation, transportation, storage, discharging, conveying and charging) are overlapping functions for the smooth functioning of the BF process which cannot be treated in isolation since they are linked processes.

The BF charge material namely iron ore, coke and fluxes are classified as bulk solids as they bear a resemblance to discrete solid particles which are in contact with each other. These materials like any granular materials are made of interacting particles. Generally, the internal structure of these materials is quite evolutionary. According to the bulk solid material classifications given in the Tab 1, charge material for the BF is bulkily classified as a broken/ discontinuous solid.

In general, the handling of bulk solids in any form has a challenge in terms of maintenance of homogeneity. Considering particles of same physical and chemical properties as discrete components of the bulk solids, it has been shown that mixing (either intentional or unintentional) of bulk solids inevitably results in natural partitioning of components. This concept is generally known as segregation (de-mixing). Naturally and / or industrially, segregation among discrete components of the bulk solids can either be beneficial (e.g. in physical separation processes) or detrimental (e.g. in mixing processes) depending on the intended function of the product or sub-unit operation.

Theoretical aspects of burden material handling

The BF burden materials consist of coke, sinter, pellets, calibrated lump ore (CLO), limestone, dolomite, manganese ore, and quartzite. Because of the nature of the process, the BF process is essentially sensitive to burden material size, the distribution, and the strength. The distribution of the burden materials in the BF affects the gas flow up movement as well as the process chemistry, heat and mass transfer between different phases in the process. The strength of the material is quite important because as the material descends, the load increases. The integrity of the burden is further compromised by subsequent chemical reactions which occur at high temperature and pressure, where attrition and breakage are prevalent.

BF coke serves as the support structure material during the BF operation as it the only burden material which descends to the hearth of the BF with partial solution reactions. It constitutes a high proportion of the hot metal (HM) production costs. In recent years there is drive to substitute coke with cheap alternative carbon sources. The normally used alternative is a direct injection of coal in the raceways in a process generally known as pulverized coal injection (PCI). However, there is a theoretical limit to the extent of the replacement as the BF devoid of adequate coke suffers from reduced burden permeability, a position which leads to the furnace choking and hanging, a scenario which further results in a furnace slipping occurrence.

The composition of the BF burden determines in-furnace process variables such as melting temperatures, softening as well as some reduction parameters, which adversely affects the production if the burden distribution is not optimized. With such limitations and challenges, there exists a need for appropriate control of the burden distribution. The typical size range of the BF burden materials is given in Tab 2.

The physical phenomena and flow structure of the granular flow, BF charge material included, appear simple on first consideration but in reality, they show a complex behaviour which is difficult to understand and predict. The situation is further complicated by the lack of direct information about the mixing and segregation parameters for such processes. However, due to the size of BF charge material aggregates and large material size distribution, the segregation tendency is a serious operational problem which is required to be reduced as much as possible.

Bulk materials handling and flow behaviour

The fundamental understanding of bulk materials behaviour and flow remains inadequate in spite of the fact that elements of powder mechanics have long been known as far back as mid-19th century. This is mainly attributable to the unique and complex characteristics of the flow physics. Handling of bulk materials shows interesting behaviour. One aspect is the ability of a macroscopic mixture of particles de-mixing due to individual particle properties. These observations are mainly as a result of the natural tendency of these materials to pattern development and self-organization. This phenomenon is mainly borne out of the behaviour of bulk materials to resemble fluids-like characteristics.

In as much as bulk materials resemble fluid-like characteristics and in itself more look-alike a solid, the overall behaviour and characteristics are completely in parallel to the observations in these common forms of matter. Ordinarily, unlike other forms of matter where the dynamics are affected by ordinary temperature, that effect is negligible in bulk materials dynamics. It is helpful to know that competing for chaotic advection or mixing effects typical of fluids is responsible for the self-organization tendency observed in bulk materials, while flow-induced segregation has no fluid phenomena.

The macro behaviour of the fundamental processes is mainly governed by micro inter-particle contact and frictional forces. Considering BF charge material sizes which are higher than 5 mm, the effects of surface forces namely electrostatic, van der Waals and capillary effects are negligible. Further, the flow behaviour of the burden materials corresponds to non-cohesive solids since cohesive tendencies are typical to superfine or ultra fine powder material of particle size less than 10 micro meters.

Classification of bulk materials flow

The flow of bulk materials can be classified according to its components, which can be defined as a group of particles with the same physical characteristics such as particle size, density, and shape. The flow structure is often difficult to analyze and is significantly influenced by particle to particle interactions as well as external excitations and boundary conditions. As a result, there is no encompassing and generic method to fully describe all bulk materials flow structure. Depending on the segment of time for which the particles are in contact, several different types of distinguished flow structures can be established.

In case of bulk material flow, the granular flow shows a variable number of metastable unsteady states. These metastable unsteady states last indefinitely if there are no external disturbances such as vibrations. The sustenance of such states largely depends on the segment of time the particles and boundaries are in contact. This depends on the solids volume fraction relative to the total volume of fluid (gas or liquid). As a result, the flow of granular materials is generally classified as dilute or dense (contact-dominated), depending on the solids volume fraction. Depending on a segment of time in contact, dense flow can be sub-categorized as either collision-dominated or contact-dominated. In these three classifications, the behaviour and characteristic flow are distinctive to a particular classification. Fluid-particle interactions (i.e. lift and drag forces) dominates the particle motion in dilute flows whereas particle to particle or particle to wall collisions or continuous particle to particle contact dominates dense flows.

Schematic representation of bulk material classification between dilute flows, collision-dominated dense flows and contact-dominated dense flows is shown in Fig 2. The given dilute bulk material classifications in the Fig 2 can be observed in cyclone separators, fluidized bed and BF material hopper flow respectively. The dense (collision and contact-dominated) flows are typical to BF bulk materials during discharge, storage and transportation processes.

Fig 2 Classification of bulk materials flow

Two classifications of dense materials flow

The shared feature of the two dense flow regimes is a structural progression as a function of length and time scales. Typically, this leads to pattern formation due to the reorientation of the particles. Such orientation of materials is driven by intermediate scale structures. The individual particle scale, better known as the micro-scale, is separated from the macro or continuum scale by the intermediate scale. Understanding complex particles interactions of granular flow assemblies are critical in quantifying segregation or de-mixing phenomenon. In this regards, the relationship between macro-structural (bulk) behaviour to the underlying micro-structural (discrete) dynamics is to be established.

Collision-dominated flows – In collision-dominated flows, the flow is dispersed, scattered and energetic, with particles interacting predominantly by near-instantaneous and binary collisions. Inertial effects can be ignored. The collisions are inelastic and energy is dissipated during particle to particle or particle to boundary interactions. Since the collisions are dissipative, a source of work is somewhat required to sustain the ‘fluidity’ of the granular material. Due to the dissipative nature of the collisions, distinctive flow behaviours and patterns such as clustering and density waves are observed. Density wave is a phenomenon linked to granular temperature, where particles do not flow uniformly but rather into regions with different velocity from that of the average velocity.

Contact-dominated flows – In contact-dominated flows, particle to particle collisions are strongly correlated, neither binary nor instantaneous but rather enduring and multiple. This flow shows two interesting characteristic scenarios where there is critical shear stress below which flow is possible and an intricate dependence on the shear rate when flowing begins. As a result of such dependence, a complete flow structure requires the inclusion of visco-plastic features associated with the contacted-dominated flow.

There are several constitutive laws proposed based on different treatments and several considerations such as adding a degree of freedom through local rotation, introduction of stochastic flow rule and modification of kinetic theory transport coefficients such as viscous terms, collision frequency, and dissipation terms. A recent formulation in which a parameter called inertia number appears to be a robust formulation, capable of reproducing a wide range of visco-plastic characteristics of the contact-dominated flows. The inertia number is the shear rate multiplied by the square-root of the particle mass divided by the pressure.

Granular materials mixing and segregation

Granular material handling is quite complicated especially when homogeneity and uniformity distribution is needed due to the natural tendency to segregation (de-mixing), especially if the sizes of the materials vary largely. Unlike fluids processes where mixing promotes homogeneity, granular flow resembles mixing and segregation mirror image in which the prolonged mixing tendency promotes de-mixing (segregation). As such, the case of BF burden materials which goes through repetitive handling processes, it is useful to treat mixing as a silent feature.

The natural order of granular material is segregating or deviation from uniform behaviour. Depending on the state of the system (charging, storage, discharging and transportation) in which granular material is exposed to, there are different forms of mechanisms by which the segregation occurs. The main drivers of segregation are a particle size difference, density difference as well as micro-properties differences such as frictional effects, the coefficient of restitution and angle of repose. Amongst these drivers, particle size seems to be the most important factor determining segregating behaviour of granular particles.

Granular material segregation mechanisms

In a study for minimizing segregation, thirteen mechanisms of granular material segregation have been proposed. However, most of these mechanisms are special or overlapping cases of other mechanisms. Considering this and to have flexibility, the classification granular material segregation has been simplified into five main mechanisms. In these simplified five mechanisms, fluidization and agglomeration segregation mechanisms refer to fine and cohesive particles respectively and are not applicable for the BF charge materials. As such, the remaining three main segregation mechanisms which are applicable to BF charge material charging, storage, discharging and transportation are described below.

Percolation or dynamic sieving mechanism – When particles with a size distribution range are made to interact (e.g. in a material storage bin) spontaneous consolidation trickling can occur with the smaller particles percolating and sieving downwards through large fraction interstices. Naturally, when granular materials flow, the granular gap widens and when this happens, interstices are created. Small particles can squeeze into small interstices below a large particle, but the reverse is much less likely to occur resulting in a net segregating flux of the smaller particles downward, away from the free surface. This action is normally known as percolation. At steady state, the particles segregate with a bottom layer having many fine particles. The percolation mechanism is shown in Fig 3. In a binary system with particles diameters, d1 and d2, where d2 is higher than d1, spontaneous sieving occurs when d1/d2 is less than or equal to 0.1547. The first use of this critical ratio was done for deriving a detailed model for segregation due to particle size differences in a chute flow based on percolation mechanism.

Fig 3 Schematic of percolation segregation mechanism

The percolation segregation mechanism is intensified by shear induction through vibration movement such as vibro-screens with even large size and high-density particles migrating to the top. This suggests that only geometrical considerations are necessary for segregation in vibrated systems. In short, for the percolation segregation mechanism to occur, three conditions need to be mainly satisfied which are the (i) critical size ratios for sieving, (ii) small particles need to be non-cohesive to pass through interstices, and (iii) existence of enough strain or inter-particle motion to increase the probability of fine particles orientation to the multiple interstices.

Trajectory mechanism – Trajectory driven segregation mechanism is normally seen at the charge material transfer points mainly filling into storage bins, discharging from the storage hoppers, and discharging at the end of conveyor belts as shown in Fig 4. In a study investigating flow conditions to ensure smooth flow of granular materials, it has been pointed out that different size materials move at different velocities due to different frictional drag which causes material segregation. The mechanism for trajectory segregation has been mathematically explained considering particles projected horizontally into a fluid. Considering a small particle of diameter d, density Dp, and drag governed by Stokes’ law, it has been shown that the maximum distance X, travelled by the particle with initial projection velocity Vi, into a fluid of viscosity Vf, and density Df, is as given in Equation X = Dp.Vi.(Dp)2/18 Vf. However, when considering BF charge materials sizes and assuming the drag force is negligible, the trajectory of the particles is given by a dimensionless trajectory equation.  The dimensionless trajectory equation is y/x = tan A – [g/2 (cos A)2]. x/Vi, where x and y are horizontal and vertical spatial coordinates respectively, A is the inclination angle, g is the acceleration due to the gravity and Vi is the initial velocity at free flight. From the equation, it can be seen that the trajectory of the particle is independent of size (mass) but rather the random velocity at the commencement of the free fall associated with each individual particle.

Fig 4 Schematic of trajectory segregation mechanism

Micro-property difference mechanism – Under this mechanism, there are three effects namely (i) frictional effect (ii) coefficient of restitution effect, and (iii) angle of repose effect.

The study of the frictional effect on granular material has shown that the granular material with two components having a different coefficient of static friction F, normally segregate. During the study, a case has been considered for a spherical particle with mass m at a height h, the coefficient of sliding friction Fs, the coefficient of rolling friction Fr, and radius r, discharging over a slope with an inclination angle A, as shown in Fig 5. Considering conservation of energy principles, if a particle at unstable equilibrium satisfying the momentum equation given in equation m.g.sin A.r = m.g.cos A. Fr. In this equation, the particle travel distance, X can be obtained by equations X = (h/Fs).(Cos A)2.(1-Fs/tan A) and X = (r.h/Fr).(cos A)2. (1- Fr/r.tan A). These two equations give the derived particle travel distances considering sliding and rolling coefficient of friction respectively. From these equations, it can be seen that particle travel distance is dependent on the particle diameter when the coefficient of rolling friction is considered and this results into segregation. On the contrary, sliding frictional effects show independence of both particle size and mass.

Fig 5 Schematic of particle motion on an inclined boundary

The mechanism of the coefficient of restitution effect is in principle a dynamic effect. When particles collide or particles impacts system boundaries, they bounce out at different velocities and final position is determined by the interfacial resilience. Example of the mechanism is charging of particles on a heap or into a storage bin. On impact at the top spot of the heap, particles with little resilient do not bounce out but get trapped at their position. On the contrary, those particles with high resilience bounce and finds a final position somewhere away from the deposition spot and possibly concentrate at the peripheral of the heap.

In case of the angle of repose effect, the mechanism of the segregation is observed during the piling of granular material. When building up a heap, it is seen that the slope angle (angle of repose) depends on the type of material and is independent of the number of particles. In a study on boundary effects on the angle of repose in rotating drums, it has been pointed out that the axial segregation is influenced by either static or dynamic angle of repose. In principle, the angle of repose effect depends on the material parameters such as particle size, distribution, shape and frictional forces and as such, it becomes a mixed mechanism due to the inclusion of other variables for the effect.

Classification of granular material segregation

Segregation of granular materials is classified based on the variables considered for the process being considered. Primarily, it is classified based on (i) physical properties of particles (i.e. size, density, or shape segregation), (ii) energy input (i.e. vibration, gravity, or shear segregation), (iii) particle movement orientation (i.e. horizontal, or vertical segregation), and (iv) equipment used (i.e. hopper, chute, or conveyor).

However, the natural or established process better known as the mechanism by which the segregation is created appears to be the most common classification approach. Segregation is widely regarded as a surface phenomenon. As such, different mechanisms observed are independent of the particles below the surface layers. In principle, the mechanisms can be accounted for by only considering the behaviour of moving surface layer particles. In various types of segregation, in principle and in most cases, overall segregation mechanism is a combination of several interacting mechanisms. For example, sieving segregation can be considered a special case of percolation and displacement/ migration segregations since they both share the principle of small particles filtering down relative to large particles.

Segregation quantification

Though there is some understanding of the mechanisms and effects of segregation on granular material flow, the issues arising from this phenomena to be adequately capped, there is a need to shift from the avoidance of the occurrence towards the control of segregation. In this regard, precise qualitative and quantitative segregation measurement methods are required to completely understand and control the effects, particularly in the case of BF burden materials, which undergo a repetitive handling process. A complete description of the particle distribution is a non-trivial exercise under such circumstances. However, in order to precisely burden the furnace stock-line, it is imperative to have the knowledge of particle material distribution along with the quantification of mixing and segregation in granular flows.

Segregation measurement indexes – In principle, segregation is a fraction of a complex granular material flow. Further, the opaqueness of granular flow makes it practically difficult if not impossible to physically extract useful data during the segregation studies. Suitable incorporation of theoretical formulations into mathematical modelling simulations can make it easier to better characterize granular flow mixing and segregation of such systems. Quantification of segregation is an important task since it involves a complete inclusion of inter-relationships between factors causing segregation and the equations of motion of the particles. However, a number of indices for measuring the quality of granular mixing, which can be used as a quantitative measure of segregation, have been proposed and they are generally expressed in statistical terms or dimensionless number terms. The common mixing and segregation metrics used is the relative standard deviation (RSD) which only accounts for the single component system. This RSD segregation metric is a reflective industrial application metric as it provides the macro mixed state of mixing for the entire system.

Mixing and segregation measurement metrics are the backbone of many indexes being used. It is important to understand that lumped sample and measure of particular variance is used to define mixing and/ or segregation index. Although the quantification is quite informative, the major drawback of such characterization is the averaging over a domain of measurement while neglecting the particle to particle distribution variation which remains undiscovered. Moreover, in the classical and fundamental concepts which have been developed about mixing and segregation of granular materials, the process of granular segregation, has been defined and concluded to be a surface phenomenon.

A number of studies have shown and proved that different shapes of handling equipment, mode of operation, and material properties can be correlated to the mixing and segregation behaviour of the material. The major challenge which still remains is to have a unifying characterization methodology, which not only describes the mixing dynamics but also essentially tackles the more complicated segregation phenomena. A new approach is to characterize the evolution of particle to particle relationships in raw materials within the charging system so as to inform the extent of mixing and segregation in time and space. The pivot of the method is based on the idea that some aspects of the relationship between particle and its nearest neighbours can be used to deduce useful insights into the evolving particle dynamics process.

Granular flow mechanistic segregation models

It has been seen that segregation kinetics is widely and in general framed into a mathematical statistics and probability framework. The apparent limitation to this approach is that the absolute reflection of the physical nature of the process is precluded coupled with the failure to prescribe the direction in which segregation is taking place. This limits the possibility of a generalization since the knowledge is quite empirical. In the first pioneering study to develop mechanistic models which incorporate all the physics surrounding the prevalent de-mixing tendencies that occur in real granular flow system, the application of kinetic theories for mixtures of granular materials has been applied to study segregation tendencies based on percolation mechanism. Using a combination of statistical and dimensional analysis, the developed formulations hold for negligible enduring frictional contacts with shear rates sufficiently high so that the dominant contributions to the total stresses are due to particle to -particle and particle to boundary collisions. In this study, it has been observed that in a chute flow with high solids volume fractions, there is a high probability of small voids formation relative to big ones. The resulting effect of such a postulation is that small particles sieves through and collects at the base of the bed. This results in a net segregation flux in a direction normal to the chute surface of the small particles.

In as much as the mechanistic models described above give some intended physical appreciation in segregation description, evaluation of key fundamentals such as dispersion coefficients of such granular flows are not small and cannot be established by the above models. In this direction, one study suggested mixed statistical and mechanical interactions based on the kinetic theory of dense gas systems since they give a general understanding of causes of granular flow segregation.

Clustering occurs as granular flow experiences damping as energy is lost after collisions. The change in velocity and movement is non-uniform hence the clusters are formed. Hence, constitutive equations have been proposed based on a kinetic theory for collisional rapid flows. The utilization of the kinetic theory expressions for the analysis of granular segregation shows that it can be used only for inelastic and different sized particles at low volume fractions. This is a limitation as typical granular flow systems are contact-dominated flows with high solids volume fractions. The application of the theory is more useful in case there are established constitutive equations for inelastic, different sized particles and high solids volume fractions.

It is seen that the granular flow resembles mixing and segregation mirror image in which the prolonged mixing tendency promotes segregation. The concentration gradients results in mixing whereas the individual contributions of pressure and temperature gradients produce segregation in granular flow systems characterized by particles with particle size range distributions and material density differences.

The theoretical aspects of the bulk handling of the materials have given a general but compact back ground on granular flow, free surface segregation, mechanisms and theoretical approaches in granular material processes. BF burden material storage, handling and transportation processes are susceptible to the fate associated with segregation. For example, the BF sinter material is known to have more in-bin size segregation and more out of bin size variation than the BF coke.

BF charging system

The charging system of the BF iron-making process can generally be described as a network of equipment and mechanisms designed to charge materials into the BF in a certain sequence, quantity and at a rate which ensure that the specified furnace productivity and prescribed stock-line level is maintained. The charging system consists of three sub-systems which are essentially responsible for (i) batching (ii) transportation, and (iii) charging into the furnace respectively. Batching is done in the stock-house which receives the bulk solid feed materials from their various sources (stockpiles, sinter plant, and coke ovens), storing each material in individual bins to provide several hours of feed material for usual BF operations. The batching process includes screens, weigh-hoppers, conveyor belts, feeders and control systems to prepare batches of charge materials. Transportation provides the means for the delivery of the materials to the top of the furnace. Normally, this is done with either by the belt conveyor system or the skip hoist arrangement. The third sub-system consists of a network of equipment and mechanisms for the charging and control equipment. The overall charging system is interconnected and controlled by an automated charge programme.

Under some conditions, furnace productivity can be limited due to the capacity of the batching (stock-house) process to deliver charge materials. This occurrence is mainly due to transient charge materials flow, equipment settings and charge requirement (burden ratios). A typical source of transient change in charge composition is caused by changes in materials delivered to the stock-house bins and is usually referred to as ‘stock transitions’. This normally occurs when the reclaimed material is used such as the substitution of fresh coke and sinter with stored coke and stored sinter respectively. Such reclaimed materials are known to alter furnace performance compared with the fresh materials. Hence, there is need and usefulness of knowing the different materials and to have their accurate tracking through the charging system so that burdening and blast parameters can be controlled optimally to maintain furnace operational stability.

One other important feature of the charging system is the mixing and segregation of charge material. For example, accurate weighing of several materials in the same hopper requires sequential delivery of the material. However, when the weigh hopper discharges, the materials inherently intermingle to some extent, yielding a time-varying composition of the delivered stream. It is imperative to have an accuracy of time-varying composition in order to estimate the radial variation in burden chemistry and physical properties of the material delivered to the furnace.

Since the burden materials undergo multi-stage handling, hence the processing of different types of charge materials need greater control for high productivity and stable operation of the BF. Also, charge material batching and transportation phenomena are required to be the key focus area for BF operator. For smooth BF operation, the operator is to be position to accurately track the burden materials upto their delivery to the furnace charging system.

The overall charging system is interconnected and controlled by an automated charge program which is coordinated by discrete event processes. Previously, BFs were generally small compared to the modern-day large capacity BFs. In small furnaces, the theoretical amount of coke was normally determined as the controlling charging factor as such, with skip charged furnaces, the optimum charging capacity is reached with full skips of coke.

In the modern BF operations, over and above the need to cope with burden material requirements of larger BF capacity, there are two additional operational factors which are (i) sustenance, and (ii) environmentally friendly operations. Sustenance is mainly through the realization of high furnace productivity which currently has been achieved by an array of technology uses. With this added dimension, the total skip weight is now normally the controlling charging factor and thus modern furnace can work with full skips of iron-bearing burden component. Considering the large size of the present day BFs, the required skip capacities become extremely large and as a result, the design and installation of such skip charging facility to cater for such a huge continuously charging system pose a challenge. Such commensuration of modern large furnaces can only be achieved with sufficient burden delivery capacity. As a result, the modern furnaces are equipped with the conveyor belts charging system. The modern BFs charging facilities consist of a stock-house with a conveyor belt transportation of burden materials to the BF bell-less top (BLT) charging system.

BLT charging system

The charge material is conveyed to the BLT charging system where it is eventually charged into BF top material hoppers (bins) which are alternately used. While one hopper is being filled, the other one is being discharged. The operation of weigh-hoppers and material hoppers is essentially the same and thus, the further description gives an account of particle behaviour during conveyance (conveyor belt), intermediate storage (material bins and weigh hoppers) and eventual discharge (chute or free fall).

Conveyor belt particles behaviour – It is seen that the granular material of varying size fractions and density cause segregation. The detailed phenomenon of the transport mechanism of granular material on a conveyor belt remains limited. However, segregation phenomena on a conveying system are difficult to explain without elaborate simplification of the problem. The system under study has to be defined in terms of mass flow rate and the conveyor speed which promote particle bed development. Operative mechanism of segregation can be established only if the system is well defined. In Fig 6, a schematic representation of the particle size segregation at the transfer point of a moving conveyor belt is shown.

Fig 6 Schematic representation of the particle segregation at the transfer point of a moving conveyor belt

It has been established that there are four main mechanisms to be considered in conveyor belt material movement segregation namely (i) percolation, (ii) particle migration, (iii) trajectory, and (iv) free surface segregation.

Material bins and weigh-hoppers particles behaviour

Granular material bins and weigh hoppers are often used for storage and eventual discharge of material to the subsequent process step. They both in principle have (i) a form of defined material feeding or filling mechanism, (ii) some retention time of material, and (iii) a defined discharge region below. All the three steps have a contribution to the overall material flow behaviour at discharge. Physical and numerical simulations have been done to clarify the relevant information about particle segregation in different kind of hoppers namely cylindrical, bins, conical, and wedge-shaped. The desirable operation is a proportionate outflow from these devices. However, since the flow is gravity induced, the outflow is not easily controlled and there are an inherent induced shear and dynamic effects which cause segregation.

The main prevalent mechanisms of segregation in material storage bins and hoppers are free surface (during feeding), percolation (during retention) and trajectory (during discharging). There is also the importance of particle size and boundary geometry dimensions during the emptying and discharge phases.

In a study to investigate how the internal angle of hoppers affect the granular flow, it has been identified some significant hindrance to free-flow for cohesion-less solids using digital particle image velocimetry (PIV) measurements. As a rule of thumb, to avoid mechanical arching (particle interlocking), the ratio Do/dp (max) is to be satisfied in the range of 5 to 10. Here Do is the boundary outlet diameter and dp (max) is a suitable maximum particle diameter. The ratio is the dimensionless characteristic scale number and it is mostly influenced by the angle of repose as well as the particle size distribution of the material.

In another study, it has been suggested that at least eight elements are to fit across the total width of any granular material handling devices in order to capture accurately the material flow rheology. This means that the diameter of the largest particle fractions in physical or theoretical experimentation is to be at most an eighth the width of a hopper, conveyor belt or any other granular material handling device outlet.

Chute flow particles behaviour

Granular material chute flow is a common feature of stock-house and BLT charging system. With the BLT charging system which comprises of the charge receiving system, material hoppers and rotating chute (distributor), chute flow has assumed additional importance. However, the core principles of the chute flow in the BF top charging system are the same as the one in the stock-house.

Chute flow can be characterized by defining three steps which are (i) burden movement before the chute, (ii) on the chute, and (iii) after the chute, as shown schematically in Fig 7. These three steps constitute three different flow classifications and as such, different considerations need to be employed to study the flow behaviour in this system. When considering burden movement before the chute, any particle collisions in this region can be ignored due to the dilute nature of the flow. When burden material is on the chute, a mathematical description can be used with velocity component along the chute being used as the initial velocity of the material flow. At the chute tip, the trajectory of the materials determines the impact point which in turn the final scatter and distribution of the material in the subsequent handling boundary/ equipment. It is possible that the mechanisms of segregation postulated for conveyor belts systems also apply to chute flow as such and segregation shown schematically in Fig 7 is possible. Three flow streams can be identified with the core flow sandwiched between lower and surface flow. At this stage, the main force considered is gravity.

Fig 7 Schematic of granular material flow in a chute

The knowledge of segregation associated with charge material is useful for understanding the charge proportioning in addressing one of the aspects of BF process intensification. However, process intensification in BF processing requires an optimized charging system capacity as BF productivity can be limited by the capacity of stock-house to supply the charge. There is a need to address and optimize multiple-handling operation stages in the product chain.

Charging system capacity analysis

The operation of the BF charging system is as critical as the design of the BF. As can be seen in the schematic representation of a typical modern BF charging system equipped with a conveyor belt delivery system in Fig 8, the material flow sequence is quite complex.

Fig 8 Typical charging system of a modern blast furnace with charging conveyor

In the interest of high productivity, the design of a BF charging system require attention to operating flexibility, availability of extra charging capability, high screening efficiency as well as a limited number of filling, discharge and transfer operations as these  cause segregation problems. One important route to increase the efficiency of the BF is full utilization of the charging system capability. Further to this, if the stock-house is not adequately designed and optimal burden delivery is not achieved, the starvation of the BF take place due to the non-availability of the burden materials which consequently results into the loss of BF productivity.

As seen in Fig 8, there are numerous unit operations in a stock-house assembly and all of them have a cascading effect on the overall performance and output delivery to the BLT charging system. In order to understand the macro-behaviour of the burden movement and overall system performance, effective and comprehensive representation of salient system elements and their relationships are to be established. Technically, this involves a description of the various handling steps, materials requirements, duration and sequencing of operations. However, for complex systems such as the stock-house, it is a huge task to clarify all the unit process information. A blend of engineering judgement, experience from similar processes, and reasonable assumptions are used for model development input data and the stock house design.

Modelling of BF charging system optimization

BF charging system involves multiple-handling material movement. The major challenge associated with multiple-handling during materials movement is the timely fulfilling of the requirement and sudden change in the process. Simulations are often used to optimize materials handling systems. Such systems generally use computer-aided process design simulators. These simulators are generally designed to model transient and continuous processes and as such they cannot be used for BF charging system operations which is a batch and semi-continuous process at best. Two available options for modelling batch and semi-continuous processes like BF charging system are spreadsheets (Microsoft Excel) and discrete event simulation (DES).

Spreadsheet models are a common platform that focuses on material balances, equipment sizing and cost analysis. Typically, the development of such a model involves writing an extensive code (in the form of macros and subroutines) in visual basic for applications (VBA) which are incorporated in Microsoft Excel. They are easy to build, much applicable to simple systems but they lack robustness and become unwieldy for large and complex systems. DES is a mathematical/ logical model of a physical system which portrays state changes at precise points in simulated time. Both, the nature of the state change and the time at which the change occurs, mandate precise description. The main feature for a successful DES is an upfront requirement of precise details regarding system and interruptions. Typically, a DES can statistically account for downtime and events. Also, modules can be created and reused while multiple grades or change in process input can be easily handled. Hence, generally the DES-based model is normally used for the BF charging system.

Cyclograms analysis is a modest DES modelling technique which has been often used in BF charging systems for its optimization. The concept evolves on the minimization function of overall start – end sequence (delivery time) of a charging cycle. The delivery time is determined by the order of activation of the mechanisms, the duration of their sequence and the length of the intervals between individual operations. It is easy to follow that the minimum cycle duration occurs when the system is devoid of pauses between the operating cycles of individual mechanisms, as well as when the mechanisms are activated in an efficient sequence.

With cyclograms analysis, it is difficult to incorporate real-time changes in system input conditions. Furthermore, the structure of the analysis precludes detailed in-cooperation of micro-system variables such as discharge behaviour and segregation tendency of materials. Due to these weaknesses of the cyclograms analysis, a DES charge material delivery model based on a mathematical/ logical representation is the better choice for the BF stock house optimization