Automation and Control System of Sinter Plant


Automation and Control System of Sinter Plant

Sintering process (Fig 1) is a metallurgical process carried out on a sintering machine. It is a thermal agglomeration process. The sintering process is an energy intensive process, in which a number of parameters have to be taken into account. The process is complex and involves various physical and chemical phenomena such as heat, mass, and momentum transfer coupled with chemical reactions. These phenomena take place simultaneously which increases considerably the complexity of the  process. The complexity of sintering process results in the complexity of its control system.

Fig 1 Sintering process

Sintering process is basically a pre-treatment process step during ironmaking which is used to agglomerate a mix of iron ore fines, return fines, fluxes, and coke breeze, with a particle size of less than 10 mm, so that the resulting sinter, with a screened size of 5 mm to 30 mm, can withstand the pressure and the temperature conditions in the blast furnace. The agglomeration in the sintering process is achieved through combustion. In this process air is sucked at the sinter strand through a bed of sinter mix. The fuel particles on the top surface layer are first ignited in a furnace and as the strand move forward, the ignited or combustion front proceeds gradually downwards through the bed until the end is reached.

Sintering process is an essential step in the blast furnace charge preparation where the agglomerate material (sinter) is required to have appropriate properties for the optimized hot metal production at the blast furnace. Further, it is also very important to control sinter plant gas emissions, which are produced in large volumes and contain a high number of polluting substances with different degrees of toxicity.

The sintering process involves a large number of parameters, more than 500, each of which exerts a greater or lesser influence and is needed to be controlled, within the possible limits, in order to optimize productivity, process stability, and standardize the composition and quality of the sinter produced. Also for meeting the statutory environmental requirements, a pollution control system is needed to monitor the particulate matter and gases generated and emitted into the atmosphere by the sinter plant. Moreover the sintering process has the characteristics of continuity, nonlinear, time-varying, and uncertainty besides complexity and large hysteresis. Further, year by year, improvements are being made to the sintering process in every sinter plant because of the accumulated experience of plant operators and the evolution and progress in sintering know-how.

For a long time, sintering process was controlled to a large extent with the experience of the operators. Due to it, fluctuations were taking place in controlling the process. Undesirable fluctuations in the control of the process were inevitable because of hysteresis, fluctuations in the data acquisition and detection of process deviations, difference in operators’ knowledge and their decision making ability, responsibility distribution among operators, and other factors such as physiological factors, psychological factors and environmental factors affected the process. The impact of these factors increased with the scaling up of the capacity of the sintering machine. This has created the necessity for the development of automation and control system for the sintering process.

Automation and control system in sinter plant is needed for meeting the above requirements. It is needed for ensuring effective control of sub systems, timely supply of the process information, and total cost minimization, while meeting the production and quality requirements. Further, automation and control system is needed for building a data base necessary for the data analysis and to incorporate the analytical tools for this purpose.

A reliable and well proven basis automation control system is the backbone of modern sinter plant operation. The main objectives of the sinter plant process control system are (i) minimizing fuel consumption since the fuel rate is a key factor in production costs, (ii) avoidance of heavy control actions since if only minor control actions are necessary, the sinter machine performance is stabilized considerably, (iii) avoidance of critical process situations because the sooner the system reacts to critical process situations, such as an inhomogeneous mixture, poor surface ignition, or incomplete burn through of the sinter mix, the smoother the overall sintering process is, resulting in a more uniform product quality, (iv) coordinated operational decisions throughout all the shifts result into constant operating conditions throughout all the shifts which increase the lifetime of the equipment and reduce production costs, and (v) reduction of emissions since with the closed-loop operation mode of the process control system, the production parameters can be optimized within the environmental emission limits, in particular, SO2 emissions.

The automation and control system of the sinter plant is a modern user friendly tool which helps in improving the productivity and stability of the sinter plant. It helps in improving the performance of the sinter plant by addressing to the needs of the plant. It stabilizes production of sinter, helps in reaching of the anticipated result and has an immense practical value. Its benefits include (i) high productivity since it keep the sinter plant running at peak performance while minimizing consumption of electric energy and fuel, (ii) product quality which means that the sinter maintains the chemical, physical, and mechanical properties sinter at the desired levels, (iii) reduced fuel consumption which is due to the result of the precise mixing of charged materials and the ideal control of the quantity of return fines taking into account the thermal conditions of the sinter on the strand, (iv) stable and shift-independent operation thus ensuring the efficient production, (v) easy integration of a comprehensive range of metallurgical models into the automation and control system, (vi) fast response to the demands of blast furnace, and (vii) quick amortization with the standard period can be expected to be less than one year.

In the sintering process, chemical and physical parameters such as basicity and product sizes are to satisfy pre-set target values within defined standard deviations in order to meet the quality requirements of the blast furnace. Sinter quality begins with the selection and mixing of the raw materials in the blending yard and dosing plant which are integrated in a common control model of the sinter process. The chemical properties are to be homogenized by an automatic adaptation of the raw material mix. An enhanced ‘burn- through point’ (BTP) control system which takes into account physical and chemical properties of the sinter mix is to be incorporated in the system. The system has to counteract changes caused by fluctuations, which is achieved by a closed-loop control of the process.

Sinter process optimization needs an innovative process control system which raises plant automation to a completely new level. It is to be based on a well-tested and proven basis system which ensures a high availability and efficiently combines data-acquisition, data-processing and data-visualization. A broad spectrum of raw data sources (front-end signals, amount of material charged, laboratory data, events, model results, and cost data) is to be stored over the whole plant lifetime. Specialized tools are to be provided where process information can be linked to analysis’ data and raw mix recipes.

Automation and control system of sinter plant has been developed with objectives of improving the labour productivity, improving the yield of the sinter pant, to get optimal sinter quality (physical, mechanical, chemical, and metallurgical), and to reduce the energy consumption. It is a vital system which assists the plant operators in the monitoring of each stage of the sinter production process.

Since the characteristics of sintering raw materials, such as chemical composition and grain size, have an intrinsic element to cause fluctuations, it is necessary to use statistical methods for the evaluation of the process data. This requires development of several models. Model development in the sinter plant is complicated because of the complex nature of the sinter process. However, models are needed to be developed which can be put into practical use online. Further, sinter plant operation is required to have flexibility for meeting the requirement of cost minimization through energy saving etc., while accommodating raw material fluctuations, meeting the requirements of the continuously improving process of ironmaking at the blast furnace.

Automation and control system of sinter plant ensures optimum and stable operation throughout the sintering process. It assists in increasing the productivity and lowering of the operational costs. For the purpose of ensuring optimum and stable sintering process, it is necessary to understand in-bed phenomena and steer the process towards optimum operation. The main control techniques in sintering are charge density control to achieve uniform sintering across the width of the strand and pallet speed control to maintain optimum productivity and sinter quality.

The efficiency of the automation strongly depends on the proper instrumentation. The right instruments are to be provided at the proper place and are to be fitted seamlessly for an efficient automation and control system. Further, established and efficient techniques such as server virtualization increase system flexibility and availability and help in saving hardware and maintenance costs. The latest industrial Ethernet technologies and proven hardware architecture permit a maximum performance while keeping a high IT (information technology) security levels.

With the ultimate aim of stabilizing the sintering process, increasing the productivity, and lowering of the production costs, automation and control system is needed in the sinter plant to assure optimum and stable operation throughout the sintering process. For this purpose, several efforts have been made to understand in-bed phenomena and steer the process towards optimum operation. The main control techniques in sintering are charge density control to achieve uniform sintering across the width of the strand and pallet speed control to maintain optimum productivity and sinter quality. The automation design of the sinter plant is normally divided into six basic equipment layers. Fig 2 shows hierarchy of sinter plant automation system.

Fig 2 Hierarchy of sinter plant automation and control system

Automation and control technologies for the sintering process were developed along with sinter technology so as to meet the sinter quality requirements needed for the ever improving performance needs of the blast furnace. Computers were first introduced at the sinter plant for detecting, alarming, recording, and printing requirements of the sintering process during the early 1960s in USA and Europe (mainly in France and Belgium). Later, computers were used for open and closed loop control step by step. It was in the 1970s when Japan became the centre of the development of the automation and control for the sintering process. Japanese steel organizations such as Kawasaki Steel Corporation (KSC), Sumitomo Metal Industries Corporation (SMI), Kobe Steel, Nippon Steel Corporation (NSC) and Nippon Kokan Keihin (NKK) carried out important projects in the field of automation and control systems. The following list includes some examples of the works which were carried out by the different Japanese organizations in the field of sinter plant automation and control. From 1980s artificial intelligence is being studied for the sintering process. Hence, control reliability and precision have been improved.

Sintering energy control system (SECOS) was developed by KSC. SECOS can detect and control the thermal energy level rapidly within an allowable range. Two parameters are considered by this control system namely (i) carbon quantity of the sinter mix which is burnt on the pallets (calculated through carbon balance by detecting the waste gas volume and composition), and (ii) hot zone ratio of the sinter cake cross section at the discharge end measured by a camera. Once evaluated the thermal energy level by using these two parameters, the coke blending ratio is adjusted. The implementation of this system has led to the Improvements in sinter quality and productivity.

Operation guidance system (OGS) was developed by the KSC with the objective of achieving a stable permeability of the sinter bed for achieving an optimal sinter quality. After inputting production data of sintering process, main system evaluates permeability, sinter quality, and productivity. It has two sub-systems. One of the sub-systems is used to assess permeability, while the other is used for auto-adjusting the standard value for assessment.

Other systems where extensive works were carried out were ‘sub-gate operation control’ by Kobe Steel, NKK, and KSC), ‘new BTP control on strand speed’ by NKK, ‘artificial dexterous nimble system’ (ADONIS) by NSC, ‘Kawasaki sinter automatic control’ (K-SAC) by KSC, and ‘unmanned operation system’ by NKK.

Automation and control system of sinter plant is structured in the classic levels, from Level 0 (field Level) up to Level 3 (management Level). It is a three-level hierarchical system which uses the distributed control system (DCS), centralized process computer system (PCS), and central computer systems (CCS) of the steel plant. DCS performs such functions as measuring wind velocity distribution and gas temperature distribution along the sinter strand, and also ‘direct digital control’ (DDC). PCS performs functions such as process control to optimize sinter plant operation, and information services to operators. Process models to apply higher level control strategies are integrated with special sinter plant measurement systems. CCS performs functions such as planning, managing, and data analysis of production and operation based on the general-purpose data base. The application of three-level control system improves flexibility, facilitates expansion, and raises the process reliability.

The three levels of control system consists of (i) digital control system (DCS) which is an integrated monitoring and control system that includes digital control computers, sensors, and transmission devices, giving local loop control and advanced control of the process with a standard of set values of the process computer, (ii) process computer system which has functions of collecting and processing data sent by the DCS, realizing set value control and comprehensive operation guidance of the process according to the instructions which  gives the ‘central computer system’ (CCS), and (iii) CCS which is the top level of the control system and has databases of the ironmaking system, which obtains by collecting and stocking information of the subordinate computers, and uses in the elaboration of the planning of material purchasing and production, but also uses in the production report and technical analysis of the production data. In this way, human resources are saved, operation costs are decreased, and operation management is improved. Moreover, intelligent systems make the sintering process less dependent on the experience of the operators. Fig 3 shows the automation and control system for sinter plant.

Fig 3 Automation and control system for sinter plant

Application of mathematical model for the process control of the total process is limited because of the complexity of sintering process. Hence development of models has preceded area wise in sinter plant. Area wise models resulted into modular approach for the automation and control system of the sinter plant.

The automation and control system of the sinter plant is normally is characterized by a modular system structure. In addition to basic functions such as data acquisition and set-point execution, the technological controls (main control loops) are implemented in the basic automation system. These include raw mix ratio control, raw mix feed control, moisture control, surge-hopper level control, drum feeder control, ignition hood control, exhaust gas cooler control, and sinter cooler control. The focus of these basic control functions is to assure a smooth and reliable sintering process and to enable a continued process optimization.

Flexible interfaces, modularization and modern software architecture provide the means to easily adapt and maintain the system in an ever changing environment with respect to the raw materials, operation philosophy, and connectivity to the third party systems. Along with the robust basis system, numerous interacting process models support the operators and metallurgical engineers in their daily decisions. Plant specific needs are incorporated into these metallurgical process models. End-to-end transparency in real-time through up-to-date data visualization and metallurgical process models lead to better collaboration, improves workflows and reduced errors while supporting decision-making.

Specific examples during the development of sinter plant automation and control system

KSC has developed sensors which monitor heat pattern indices in the sinter bed. These sensors allow more elaborate operation control, not only by evaluating operating conditions from the transition of permeability, quality, yield, and other conventional time-series data, but also by more directly estimating changes in bed conditions and controlling the wind volume distribution and coke distribution in the bed. Fig 4 shows an example of heat pattern achieved by measuring system at the sinter plant.

Fig 4 Heat pattern achieved by measuring system at the sinter plant

ArcelorMittal in collaboration with the Centre de Recherches Metallurgiques (Belgium) developed in 1992 an expert system with Nexpert Object software and VAX VMS hardware which controls 200 operating parameters every 15 seconds. The benefits which have been achieved include lowering the standard deviation of the FeO content, improving the RDI (reduction degradation index), improving the control of the BTP (defined as the point where the temperature of the waste gas reaches its highest value, which happen when flame front reaches the bottom of the sinter bed), and the temperature profile in the wind boxes.

In 1994 ArcelorMittal installed a sinter strand control system based on analysis of the CO (carbon mono-oxide), CO2 (carbon di-oxide), and O2 (oxygen) content in the wind box exhaust gases, coupled with the staged heat balance model developed by IRSID (Institut de recherche de la siderurgie). The experience obtained with the use of the control system has indicated variation in the productivity, BTP, maximum flame front temperature and the fraction of melt which forms in the bed at the maximum temperature, as a function of the variation of a series of parameters. The theoretical productivity depends mainly on the return fines balance and on the raw mix flow. Assessment of the difference between real and measured productivities has shown that the effect of the return fines balance is well traced.

The calculated BTP depends on the flow and the chemical composition of the raw mix (through the carbon content in the mineral mix and the moisture content in the fuel) and the CO2 in the exhaust gas. Assessment of the difference between the calculated BTP and that resulting from temperature measurements under the grate has shown variations on both sides of a mean deviation of the order of 1 m to 2 m. An increase in this value constitutes a warming. The calculated maximum flame front temperature reached at the flame front and the fraction of melt depends on heat inputs and requirements connected with the chemical composition of the raw mix.

British Steel Corporation (BSC) has installed in 1994, a VAX 4000/500 central computer for logging data from 4500 signals along with two MicroVAX 3100 Model 80s to act as auxiliary machines to drive terminals and printers. Use of the computer has improved the determination of the optimum raw mix moisture following a sinter bed change. By being able to constantly monitor operating data, it has been possible to calculate permeability on a continuous basis, based on gas and air flows into the ignition furnace. Fig 5 shows relationship between blend moisture content and the permeability. It can be seen that the optimum moisture can be easily observed. It is also possible to see the known strong influence of moisture on the permeability.

Fig 5 Relationship between blend moisture content and permeability

For the measurement of the FeO content in the sinter, a belt coil system has been placed around the product conveyor belt. Improved FeO control has allowed a coke saving of 1 kg per ton of sinter. It has also been possible to improve process control and achieve higher strand use efficiency by measuring and controlling the BTP, where the maximum temperature occurs. A BTP model has been installed on the plant computer to calculate and detect the BTP along the sinter strand by measuring the under-strand temperature for the last eight wind boxes. The model calculates a proposed strand speed to bring the actual BTP into line with a user-specified setting, and this is notified to the operator in control of the strand. When strand speed has been controlled using this model, strand utilization has increased by around 1 %.

Wuhan Iron and Steel Company (WISCO) in China in collaboration with Voest-Alpine Industrieanlagenbau (VAI) of Austria installed a model which calculates 700 values at two levels of automation and a set of process optimization modules. In spite of the great complexity of the process parameters, the main objectives have been (i) improvement of output by 5 %, (ii) reduction of coke consumption by more than 2 %, and (iii) ensuring system availability higher than 99.5%.

The first level of the model includes all the main data from different sources, such as the analysis of raw materials and sinters. This data is presented to the operator in a very efficient way, normally in graphic form. This data is analyzed by the operator in graphic format. The level 2 model provides closed-loop control of the process (without human interaction). Several set points are calculated by the model at any time and simultaneously are verified if these set points are applicable. These set points include raw mix flow, material ratios, water addition, and sinter strand speed to control the BTP. The system can switch from the level 1 to the level 2, and if the level 2 control is not possible for any reason, control is again taken over by the level 1.

When it is not possible to regulate the strand speed in an ideal way, speed variations can be expected to cause unstable operation. For controlling the process it is necessary to assure complete sintering of the mix before reaching the end of the strand, and assuring that the BTP point is as close as possible to the end of the strand in order to achieve the maximum output. Fig 6 shows overview of the sinter plant productivity control.

Fig 6 Overview of sinter plant productivity control

Pohang Iron and Steel Company (POSCO) has installed at Pohang an integrated visual monitoring and guide system in the sinter plant which consists of fifty thermocouples, 5 on-bed flow meters, a thermal imaging device, and a sub-gate opening control system. Fig 7 shows the layout of this visual monitoring system. Thermocouples are placed below wind boxes 15 to 25 and the 5 flow meters are on the bed box 6. Using this system, the operator can maintain the process in optimum and stable conditions. The visualization of the strand state gives information on the pallet speed, charging density, and fuel content to the operator. As a result, fluctuations in the off-gas temperature under the sinter pallet decrease considerably. Also, after automatic control of the sub-gate opening, and so, lower quantity of return fines is produced and higher product yield is reached.

Fig 7 Visual monitoring system layout

POSCO has installed at its Gwangyang plant an on-line measuring system to check the moisture content in the iron ores and coke sent to the drum mixer hoppers. The system is based on a fast neutron source. When the fast neutrons collide with hydrogen atoms in the mineral water, they lose energy in proportion to the number of hydrogen atoms, and the moisture content can be determined using a sensor which detects the slow neutrons resulting from the collision (Fig 8). The system is calibrated to take into account variations in the ambient temperature and humidity over the year.

Fig 8 Calibration line compensation system

Development of models for sintering process

During the process of sintering, several chemical reactions and phase transformations take place, not only due to the heat front changes, but also due to the modifications of local gas composition and initial melting temperatures of the mixture of raw materials. When local temperature and composition of the solids is reached, mostly the phase transformations are driven by heat supply and diffusion which take place within the particles bed with the mechanism of liquid formation playing the major role. The materials partially melt down when the local temperature reaches the melting temperature and as it moves, the contact with cold gas promotes the re-solidification and thus, the particle agglomeration forms a continuous porous sinter cake. The final sinter cake properties are strongly dependent upon the thermal cycle, initial chemical composition of the raw materials, and the thermo-physical properties developed during sintering. The mathematical models of the sintering process simulate the phenomena taking place within the sinter machine.

The method for modelling the sintering process is based on multi-phase, and multi-component transport equations of momentum, mass, and energy for gas, solid and liquid phases taking into account the local phenomena of porous sinter formation (Fig 9). The model considers the phases interacting simultaneously and the chemical species of each phase is calculated based on the chemical species conservation equations. The accurate descriptions of rate exchange for momentum, energy, and chemical reactions are essential for the complete accuracy of the model.

Fig 9 Multi-phase transfer of momentum, mass, and energy considered for sintering model

The chemical species are individually taken into account by solving the transport equation of each chemical species of the gas and solid phases. The solid phase accounts for the mixture of iron ore sinter feed, fine sinter (returned fine sinter), coke breeze (or other solid fuel), scales (fines from steel plant), and fluxes. The liquid phase is composed of melted and formed components in the liquid phase. The re-solidified phase comprises the liquids re-solidified and phases formed during the re-solidification process and strongly depend on the local liquid composition and heat exchange. The final sinter cake is formed by a mixture of these materials and its quality depends upon the final compositions and volume fractions of each of these materials and their distribution within the mosaic sinter structure.

In the sinter process model, it is assumed that the liquid phase formed moves together with the remaining solid phase due to the viscosity and considering that the liquid is formed attached on the surface of the unmelted particles, thus, equations for momentum transfer and enthalpy of the solids account for this mixture of viscous liquid and solid materials. In the model, the temperature-composition dependent thermo-physical properties are assumed to follow the mixture rule to take into account the individual phase properties considered by their phase volume fractions.

During 2012, a prediction model was developed by Hauck et al for the wind box temperature profile and burn-through point position and temperature, which was integrated into a process stabilization control system. In this way, stabilization of the quality parameters, increasing of the productivity and reduction in the fuel dosing were achieved.

In 2012, a computational simulation of the sinter process has been developed which was able to predict the most important phenomena within the sintering bed. The model was based on the multi phase concept with multiple components described by conservation equations of each component coupled with the momentum, chemical reactions and heat transfer. The model validation was carried out comparing the model predictions with averaged industrial data and local temperature measurements within the sinter strand. The model predictions presented good agreement with the averaged values measured on the industrial sinter process.

In 2013 a sinter mathematical model was developed by Saiz and Posada with the objective of controlling the BTP in sintering plants (position and temperature). Stable BTP leads to stable sintering process and the improvement of both quality and productivity. This mathematical model has been applied to the sinter plant of ArcelorMittal at Asturias (strand speed and coke consumption as control variables, BTP position, and temperature have been used in the control algorithm). By applying the model, an improvement in the productivity of 4 % to 5 %, reduction in coke consumption from 5.2 % to 5.5 % and decrease in return fines from 37 % to 45 % have been achieved.

Fuzzy logic allows for the coordination among the objective knowledge (e.g. mathematical models) and the subjective knowledge (e.g. linguistic information which cannot be quantified by means of the traditional mathematics, as the plant operators’ information). In this way, fuzzy logic has been recently used. A new approach based on fuzzy inference to control the charging gates of the sinter plant has been studied in 2014. Two strategies have been established. One which is more invasive within process operational conditions, used in sinter plants with low productivity (for example when the plant is restarted after a stoppage) with the purpose of maximizing the productivity and the second which is more conservative, also aims to maximize the productivity but the machine shows high sensitivity to changes on the gates.

During 2016, software for sinter cost optimization based on the sinter quality characteristics has been developed. Moreover, Tumbler and RDI indexes have been estimated by means of a Sugeno-type fuzzy inference system. Historical data of 6 years has been used in the multivariate statistics studies, and thus, software which gives the least expensive blend mix to be used in the sintering process and satisfies the quality requirements has been developed. Economic savings have been achieved and sinter quality has improved.

Modelling and simulation knowledge has allowed the possibility of evaluating the effect of different variables in processes, concretely in the sintering process without using directly the real sinter plant. The effect of coke combustion rate on the temperature distribution in iron ore sintering process and the growth of voids and cracks in the sinter cake by using simulation techniques have been studies. During 2015, the recovery of waste heat from sinter cooling process has been simulated and optimized.

Leave a Comment