Actual Furnace Research Articles

Fractional control of an industrial furnace

The requirements of high production allied with product quality, process safety and environmental regulation, lead control systems to play a key role in the operation of chemical and biochemical plants. In petrochemical plants, furnaces are essential equipments for process operation and due to energy costs, adequate operation and control are of extreme importance for process economics. The search for new and more efficient control laws led to the development of fractional PID control algorithm, which is based on the use of fractional differential equations. In this work, a previously identified mathematical model of an actual industrial furnace is used for fractional PID control studies. Feedback loop in servo control was analyzed, focusing on the study of the influence of the controller parameters over control loop performance. Particularly, P, fractional PI and fractional PD controller were considered in this study. Simulations were carried out showing that the fractional controllers were able to perform set-point transitions. The control loop performance was evaluated by ITAE and ISE criteria, showing that, in this study, fractional PI is the best algorithm.

Integrated Hybrid-PSO and Fuzzy-NN Decoupling Control for Temperature of Reheating Furnace

This paper presents an integrated method of intelligent decoupling control as a solution to the problem of adjusting the zone temperatures in a regenerative pusher-type reheating furnace. First, a recurrent neural network (NN) for estimating the zone temperatures and a heat transfer model for predicting billet temperatures are built based on data from actual furnace operations. Next, a decoupling strategy in combination with a fuzzy NN is used to control the zone temperatures. The architecture of the controller is based on a fuzzy c-means clustering approach; and the weights are optimized by a hybrid particle swarm optimization (HPSO) algorithm, which integrates the global optimization of density-based selection and the precise search of clonal expansion in an immune system with the fast local search of particle swarm optimization. HPSO is also used to optimize the zone temperature settings to minimize three items: fuel consumption, the temperature gradient within a billet, and the error between the mean and target temperatures of a billet at the furnace exit. The results of actual runs demonstrate the validity of this method.

Mathematical Modeling of Sulfate Fining in a Glass Melting Furnace

In this study, we propose a new mathematical model of dissolved SO3 transport in a glass melting furnace, considering secondary decomposition of sulfate. The calculated concentration of SO3 agrees with measured data near the maximum temperature area, hot spot, in an actual glass tank furnace. The effect of onset temperature of secondary decomposition, the maximum solubility of SO3 and concentration of inlet SO3 on sulfate fining is quantified by simulation.

Model based energy benchmarking for glass furnace

Energy benchmarking of processes is important for setting energy efficiency targets and planning energy management strategies. Most approaches used for energy benchmarking are based on statistical methods by comparing with a sample of existing plants. This paper presents a model based approach for benchmarking of energy intensive industrial processes and illustrates this approach for industrial glass furnaces. A simulation model for a glass furnace is developed using mass and energy balances, and heat loss equations for the different zones and empirical equations based on operating practices. The model is checked with field data from end fired industrial glass furnaces in India. The simulation model enables calculation of the energy performance of a given furnace design. The model results show the potential for improvement and the impact of different operating and design preferences on specific energy consumption. A case study for a 100 TPD end fired furnace is presented. An achievable minimum energy consumption of about 3830 kJ/kg is estimated for this furnace. The useful heat carried by glass is about 53% of the heat supplied by the fuel. Actual furnaces operating at these production scales have a potential for reduction in energy consumption of about 20–25%.

Slagging properties of coals and allowance for them in the analysis of the possibility of using out-of-design fuel

It is shown that out-of-design coal should be used only after computational analysis of slagging of the boiler. The use of ash fusibility as a criterion for the analysis should be avoided because this property does not reflect the slagging properties of coal to an acceptable degree. Parameters characterizing the properties more appropriately are presented. It is shown that slagging of an actual furnace depends on many factors and the condition that “the thermal stresses should be lower than the limiting value” is insufficient. An algorithm for the analysis is suggested on the basis of the slagging conditions arising upon transition from one kind of fuel to another with allowance for the actual slagging situation.

Blast furnace campaign life relating to the productivity

The effect of blast furnace productivity on the erosion of the hearth wall and the campaign life was examined. First, the molten iron flow in the hearth was studied by a model experiment and an iron flow estimation model calculation. Next, the temperature distribution in the hearth wall was calculated by a heat transfer model, and the erosion of the wall and formation of the solidified layer were estimated. The study led to the erosion pattern during the blast furnace campaign life, which was verified by the actual furnace erosion pattern.

The effect of combined radiation and convection on hot dip galvanizing kettle wear

Hot dip galvanizing furnaces are designed to maintain a quantity of molten zinc at 450 °C in a kettle to allow steelwork to be galvanized. The most common heating medium for such furnaces both in the UK and abroad is natural gas. The alloying reaction between the zinc and the steel kettle wall is temperature dependent. Above a certain rate of heat transfer this forms a non-adherent alloy, which allows molten zinc to erode away the kettle wall, limiting the life span of the kettle and the production rate for the furnace. In order to investigate the causes of high levels of localised wear on kettle walls a model of a hot dip galvanizing furnace has been built using CFD software. This model is compared with measurements taken from an actual furnace in operation and also from wear data of the kettle walls after the serviceable life. In contrast to received wisdom, which is that the high wear is a result of direct flame impingement onto the kettle wall, the modelling suggests that radiation and convection exchange from the flame causes the wear rather than direct contact between the flame and kettle wall.

Study of solid mushroom formation in direct iron ore smelting reduction process using low temperature water model

In the direct iron ore smelting reduction process, molten iron near the bottom blowing gas tuyere is cooled by low temperature/endothermic gas and forms a mushroom shaped solid on top of the tuyere. The formation of an appropriate solid mushroom, which covers the tuyere, can protect the tuyere and the surrounding refractory. In the present study, a water model with a low temperature gas system was established to investigate formation of the solid mushroom and the effects of operating conditions on its shape and dimensions. Transparent acrylic was used to construct the water model, which was 40% of the size of the actual furnace. Water was used to simulate the molten iron. Low temperature air, obtained by passing air through a heat exchanger cooled by liquid nitrogen, was blown into the water bath through a bottom tuyere. The air temperature was able to reach-188±1°C. In the water model experiments, water near the tuyere was cooled, and formed an ice mushroom surrounding the tuyere. The effects of operating conditions, mainly gas flowrate and mould material surrounding the tuyere, on the parameters of the solid mushroom were investigated. The parameters of the solid mushroom included whether it could be formed and duration of the solid mushroom, as well as the shape, dimensions, and weight of the solid mushroom. Attempts were also made to relate the temperature-time and pressure-time relationships of the blown gas to the parameters of the solid mushroom. With copper used as mould material surrounding the tuyere, the water model experiments were conducted with flowrate of the bottom blown gas set in the range 30-90 NL min-1. The results show that as the gas flowrate was increased, the highest water temperature which allowed the solid mushroom to form in the water model was increased. Three different types of pressure-time curve were obtained under different gas flowrates in the present study. They also corresponded to different forms of solid mushroom. As peaks appeared in the pressure-time curve, they revealed ice capsulation and subsequent bursting to release the pressure. A gas flowrate of 80 NL min-1 and water temperature of 19·2°C with copper plate as the bottom material are considered to be optimal conditions of the water model for growth of the appropriate ice mushroom. These data are rather consistent with the gas flowrate and superheat for the actual direct iron ore smelting reduction unit, which are 2700 NL min-1 and 120°C (equivalent to 70 NL min-1 and 22·7°C in the water model).

Kinetics of Silicon Transfer from Pulverized Coal Injected into Blast Furnace under Intensive Coal Injection

With the progress of shifting to operation of intensive coal injection, Si in molten metal tends to increase, and this phenomenon cannot be explained by a conventional estimation model which does not distinguish between coke and pulverized coal on contribution to Si in molten metal. With respect to SiO gasification reactions from SiO 2 in pulverized coal, basic experiments were carried out on SiO(g) generation from pulverized coal char packed in a crucible under flow of reducing gas, and reaction rate was deduced with SiO 2 concentration in pulverized coal ash and slag composition taken into account. Furthermore, with respect to Si transfer in blast furnace, the SiO(g) generation rate derived from pulverized coal under intensive coal injection was evaluated and effects of pulverized coal on Si in molten metal were investigated. Consequently the molten metal [Si] variation associated with changes of pulverized coal brand in an actual furnace is become to be estimated.

Estimation of Stack Profile of Burden at Peripheral Zone of Blast Furnace Top

We, in an off-line distribution testing equipment, have developed a device capable of measuring the stack profile with high accuracy through an ultrasonic sensor. We have further devised a method which derives distribution of inclination angle, angle of repose and terrace length quantitatively and automatically by making use of the measurement data acquired from the device. We have also invented a method estimating theoretically the stack profile at peripheral and confirmed that it accords well with the measurement results obtained from the experiment or at an actual furnace.

All Pellets Operation in Kobe No. 3 Blast Furnace under Intensive Coal Injection.

In Kobe No. 3 blast furnace (third campaign) (inner volume: 1845 m3; blown in on April 5, 1983), the use of pellets was begun as a result of suspended operation of the sintering plant in the end of May 1999 and Kobe 3BF began all pellets operation (pellet: 73%; lumpy ore: 27%) in the end of September 2001. This is dedicated to the advancement of all pellets operation under intensive coal injection with the bell-less blast furnace. A study has been made of the deepening of understanding on the phenomena of softening and melting properties by using the mixture of dolomite-fluxed pellets and low basicity pellets. In order to prevent the growth of cohesive zone root by using of low basicity pellets, the charging technique of time-series discharging based on model experiments was developed and applied to an actual furnace. To achieve the high accurate burden distribution control, the change of ore falling trajectory by fling-up of pellets inside the rotation chute was monitored and corrected. With respect to slag and metal chemistry under the low slag ratio operation with use of low basicity pellets, securing the slag fluidity and desulfurizing efficiency was designated as control points.

Control of CO2 Peak Position by Dual Lance Air Curtain Method.

A phenomenon that a CO 2 peak position or temperature peak position of combustion (CO 2 or temperature peak position) in the raceway of blast furnace approaches a tuyere in step with the increase of pulverized coal rate (PCR) changes a gas flow at the lower part of the furnace into a peripheral flow or lowers the temperature at the raceway end, which may be one of the factors making the furnace operation unstable. Therefore, a pulverized coal combustion method (dual lance air curtain method) which interrupts temporarily a contact of pulverized coal and hot air in the vicinity of lance tip by injecting cold air through an opening between inner and outer tubes of the dual lance was examined by performing a hot model experiment and a single tuyere test of actual furnace. The hot model experiment results have revealed that with the increase of the flowing velocity of cold air sent from the outer tube of dual lance, the temperature around the tuyere tip falls and the temperature ranging from the raceway and to a dead man surface rises. According to the examination results of gas temperature under the raceway conditions of actual furnace with the use of one-dimensional raceway model, an estimated value of the moving distance of CO 2 or temperature peak position of combustion toward the dead man side is 480 mm when the dual lance air curtain method has been taken this time at Kimitsu No. 2 blast furnace. The combustion efficiency of pulverized coal could decrease due to movement of CO 2 or temperature peak position of combustion to the dead man side. And in the actual furnace test conducted this time with a single tuyere, the increase range of unburnt char ratio at the inside of dead man is as small as less than 1 % relative to -1 mm fine and the packing structure of dead man surface has been improved.

High-temperature meltdown characteristics of DRI-flux mixture in a melter-gasifier

The slag formation mechanism in the melter-gasifier was examined by the meltdown test for DRI-additive-ash mixed bed under load and by analysis for the slag taken from the actual furnace of the two-step ironmaking process: For the slag phase analyses, chemical analysis, SEM-EDS and XRD pattern analysis were employed. The primary slag originated from the gangue in DRI was homogeneous. Partially-melted secondary slag consisting of dicalcium silicate with dissolved FeO was formed under rather higher CaO/SiO 2 basicity compared to that of tapped slag. Then, these slags reacted with bottom ash resulting in the homogeneous tertiary slag with the dissolved FeO. Finally, the final slag, similar to tapped slag, was collected on the top of the metal pool. The slag had CaO/SiO 2 slightly greater than 1 in the hearth of the melter-gasifier.

Dead-man temperature estimation based on the acoustic CT method

A new sensor measuring the two dimensional temperature distribution of the dead-man in a blast furnace has been developed. The principle is an acoustic wave computed tomography method based on the theory that acoustic velocity in gas depends on temperature.The features of a new measurement apparatus and the measurement results in actual blast furnaces carried out during the scheduled shutdown are presented and discussed.

Corrosion Prevention of Waterwall Tube by Field Metal Spraying in Municipal Waste Incineration Plants

Abstract Fireside corrosion factors in high-temperature corrosion of waterwall tubes in waste incineration boilers were investigated. Factors accelerating corrosion were considered to be heavy metal (zinc and lead) chlorides, unburned carbon in deposits, and reduction of the gas atmosphere formed by the increase of unburned gas components and the decrease of oxygen (O2). Laboratory tests confirmed that increasing these factors increased the corrosion rate of carbon steel (CS). These tendencies were in good agreement with those in actual furnaces. A gas spray coating of two-layer Al/80% Ni-20% Cr, selected after laboratory corrosion tests, was applied to waterwall tubes at the test site. The Al/80% Ni-20% Cr spray coating maintained excellent durability for > 3 years, and its effectiveness in corrosion prevention was confirmed. This coating system has been applied at seven plants. Main deterioration factors affecting the coating layer were considered to be formation of corrosion scale and a decrease in bon...

Radial Distribution of Burden Descent Velocity near Burden Surface in Blast Furnace.

An examination was made in regard to the influences of various factors on descent characteristic at a free surface and inner burden based on the model experiment and measurements at an actual furnace. At the same time a non-uniform descent model was prepared and a study was conducted with the use of a non-uniform descent model as to a mechanism forming the radial distribution of burden descent velocity near the burden surface in a blast furnace. As a result, the following knowledge has been acquired. The relative burden descent velocity at the free surface is relatively larger at the peripheral area and smaller at the central area than that at the inner burden. The relative burden descent velocity at the peripheral area on the free surface is large when a peripheral ore/coke is high or there is a wall erosion at the upper part of shaft. Non-uniform descent characteristic at the free surface is remarkable for alumina ball and sintered ore having a small repose angle and not remarkable for coke having a large repose angle. In the actual operation, radial distribution of relative burden descent velocity near the burden surface greatly varies depending on a relative ore/coke distribution, difference in particle size between coke and sintered ore and physical property values, even if the charging pattern is the same.

Fundamental experiment on the combustion of coal-water mixture and modeling of the process

A fundamental study of the combustion process of coal-water mixtures has been performed. The surface and center temperatures and the weight change of a single droplet under combusion were measured in detail, where the water content, properties of coal, initial diameter of droplet, and external heat flux were parameters. In the experiments, an infrared intensive laser was used as the external heat source to simulate the conditions around the droplets atomized in an actual furnace. Based on the data obtained, the reaction rates in the processes of water vaporization, devolatilization, and char combustion were formulated. For the devolatilization stage, the activation energy derived by assuming a first-order reaction was almost the same as that for the mother coal, but the apparent frequency factor was much, smaller than that for the mother coal by 1/140-1/50. For the stage of char combustion, also, the apparent frequency factor was much smaller. The specific surface area and the micropores in the coal particle were much reduced after undergoing processes of water vaporization and devolatilization. It's possible that the pore would be masked when the water trapped in the pore was escaping from it with the agglomeration of coal particles proceeding. According to the data obtained and the combustion model proposed here, computer simulation codes for the combustion of a single droplet and for the combustion process in a furnace, were written. The calculated results for the temperature variation and the concentrations of oxygen and unburned carbon in the exhaust gas agreed well with the experimental data obtained using a one-dimensional test furnace. This confirms the validity, of the combustion model and reaction data.

Natural convection induced by free surface heat flux and temperature difference between left and right walls in glass melting furnace

A numerical study on natural convection induced by free surface heat flux and cold left and hot right walls in glass melting furnaces has been performed. A function of heat flux derived from the combustion environments of actual glass melting furnace is applied to thermal boundary condition at free surface. Fundamentally there exist two flow cells in cavity (left counterclockwise one and right clockwise one). The effects of heat flux and Rayleigh number are investigated through two-dimensional steady-state assumption. The convection strength of two flow cell located in left region continuously increases. In the mean time the strength of flow cell in right region increases and then decreases. Critical Rayleigh number in which two flow cells take place above and below show linear dependence on the free surface heat flux. To maintain the traditional flow pattern (left and right flow cells) in glass melting furnace, Rayleigh number is recommended to be below 10 .

Technology of high-rate pulverized coal injection into blast furnace

Based on the results obtained from off-line model experiments and actual furnace tests, a high-rate pulverized coal injection operation was carried out at Kimitsu No. 3 blast furnace of Nippon Steel Corporation. A technology to achieve a stable blast furnace operation at the PCI rate of 190 kg/t has been established. For a stable operation at PCI rates greater than 200 kg/t and coke rates below 300 kg/t, the reduction-melting behaviour of ore at the lower part of the furnace has to be improved.

Proportional-integral-derivative-controlled numerical model for the programmable multizone furnace

This article presents a two-dimensiona l axisymmetric combined conduction and radiation model for the prototype of the programmable multizone furnace, which is under design at NASA Lewis Research Center. The actual furnace is intended for growing high-quality semiconductor crystals and performing other materials processing experiments in space. A novel feature of the model is a control algorithm that automatically adjusts the power in any number of independently controlled heaters to establish the desired sample temperatures during the numerical simulations. The control algorithm eliminates the need for numerous trial and error runs previously required to obtain the same results. The finite element code, FIDAP, was modified to directly incorporate the control algorithm. This algorithm, which presently uses proportional-integral-derivative (PID) control, and the associated heat transfer model are briefly discussed. Together, they have been used to predict the heater power distributions for a variety of furnace configurations and desired temperature profiles. Examples are included to demonstrate the effectiveness of the PID-controlled model in establishing isothermal, Bridgman, and other complicated temperature profiles in the sample. An example is also given to show how the algorithm can be used to change the desired profile with time according to a prescribed temperature-time evolution. Nomenclature A = surface area C = specific heat D = derivative coefficient E = error between actual and set point temperature F = matrix containing boundary condition information FkJ = view factor between two surfaces k and j h = heat transfer coefficient / = integral coefficient K = global system matrix containing conductivity information /Cth = thermal conductivity, scalar L = heater wire length M = matrix containing density and specific heat information