GB1574834A

GB1574834A - Monitoring a blast furnace during operation and controlling operation of the furnace accordingly

Info

Publication number: GB1574834A
Application number: GB4996876A
Authority: GB
Original assignee: Nippon Steel Corp
Current assignee: Nippon Steel Corp
Priority date: 1976-03-15
Filing date: 1976-11-30
Publication date: 1980-09-10
Also published as: FR2344631A1; DE2655297B2; DE2655297A1; AU2005576A; DE2655297C3; IT1068013B; FR2344631B1; AU507762B2

Description

(54) MONITORING A BLAST FURNACE DURING OPERATION, AND CON TROLLING OPERATION OF THE FURNACE ACCORDINGLY (71) We, NIPPON STEEL CORPORATION, of 6-3, Otemachi 2-chome, Chiyoda-ku, Tokyo, Japan, a corporation organised under the laws of Japan, do hereby declare the invention for which we pray that a patent may be- granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement: This invention is concerned with blast furnaces. More particularly, in its several aspects, it relates to a blast furnace, to a method of monitoring a blast furnace during operation, and to a method for controlling the operation of a blast furnace.

A blast furnace consists of a shaft furnace supplied with a charge of material to be heated and solid fuel from the top and having a gaseous current introduced into the lower part thereof. Such furnaces are conventionally of a substantial size, the interior being effectively sealed, and operate at high temperature within a thick layer of a refractory substance.

Consequently, it has been generally regarded as difficult to establish and follow the behaviour of the charge and gases within the furnace with any degree of accuracy.

However, in order to improve productivity of the furnace and in particular to improve the quality of pig iron produced, it is highly desirable that the conditions within the furnace (hereafter merely referred to as "furnace conditions") be maintained and controlled in a stabilised manner.

Furnace conditions may be classified in various ways and in terms of particular criteria. A convenient classification is into conditions indicative of the furnace thermal state, and conditions indicative of the movement of gas and charge through the interior of the furnace.

These latter conditions may be sub-divided into the following categories: (i) The gas permeability defined in the narrow sense, and which may be determined in terms of the gas pressure lost in the furnace or in terms of fluctuations in the blast pressure; (ii) Conditions indicative of the downward movement of the charge, particularly having regard to the shape of boundary surfaces between layers of charge and defective downward movement in terms of such phenomena as hanging, slip and the like; and (iii) Gas flow distribution in a cross section of the furnace, the distribution of velocity of downward movement of the charges within the furnace, and the distribution of the thickness of layers of charges within the furnace.

The furnace thermal states may be further sub-divided into the distribution of temperature both in vertical and horizontal directions of furnace.

Temperature in the direct reduction zone in the lower portion of the furnace, and in particular the temperature of the hot metal, is of importance, as are the Si content of the pig iron and any fluctuations therein, temperatures in the body of the furnace including the shaft section, and the temperature of the gases at the top of the furnace. The composition of the gas at the top of the furnace, particularly in terms of CO, CO2, H2 and N2, whereby the condition of reduction of ore in the shaft section and in the lower portion of the furnace may be calculated, are also considered important items regarding the furnace thermal state since they are influenced by the thermal conditions of the direct reduction.

We shall now explain further the inter-relation between the conditions indicative of the movement of gas and charge within the furnace and the conditions of furnace thermal state and in particular the role of these conditions in the operation of the furnace.

The conditions indicative of movement of gas and charge are largely determined by the physical properties of charge and the way in which the packed layers are disposed in the blast furnace. For instance, if there should be a spread in the distribution of particle size of the coke, ore, and other charged materials, or if the strength thereof should decrease a deterioration in permeability within the furnace will result. To overcome this problem, the charge is usually sieved prior to charging to avoid finely divided powder being charged into the furnace, so far as possible. The behaviour of charges fed into the furnace has come to be regarded as a matter of some importance for the operation of blast furnaces. It is believed that the distribution in the thickness of layers of coke and ore, the particular configuration adopted by the layers, and the distribution of the velocity of descent of the charges exercise a substantial influence both upon the conditions indicative of movement of gas and charge within the furnace and the conditions indicative of the furnace thermal state. Coke has a larger mean particle diameter than the ore, which is believed to reduce the resistance to permeation. Although the ore tends to soften and to fuse in a temperature of 1,000"C or more to form a fused layer having a high resistance to permeation, coke, on the other hand, maintains a virtually solid state right down to the combustion zone adjacent the tuyere.

Thus, the permeability of the furnace may well be considered to be determined by the way in which the coke is filled into the furnace, and in particular the distribution of the thickness of the layers of coke within the furnace.

It has generally been the practice in blast furnace operation to charge coke and ore into the furnace at a charging position in substantially flat alternating layers. The distribution in the thickness of the coke and ore layers are consequently inseparable one from the other and if one can be determined, the other follows. Moreover, knowledge of the distribution of these layers would be of considerable value in monitoring operation of the furnace or in effecting a degree of automatic control of the furnace. Increase of the thickness of coke layers in the central region of the furnace will enable gas to pass readily through the central region of the furnace. If the gas flow through the central region of the furnace is excessive with too great an emphasis placed on permeability, chemical reaction between CO, H2 and ore in the indirect reduction zone in the upper portion of the furnace will be diminished.

The efficiency of reduction of the ore will be decreased until the point is reached where direct endothermic reduction will increase resulting in a shortage in furnace heat. In such a case, if the thickness of the ore layer in the vicinity of the furnace wall or in intermediate radial positions is increased, a situation may result in which the calorific value required for melting ore is increased so that there will be an increase in the melting load which may lead to an undesirable cooling of the furnace.

A decrease in the thickness of the ore layer at or in the vicinity of the furnace wall results in a corresponding increase in the thickness of the coke layer in the same region alleviating the melting load in the combustion zone at the tuyere. However, this will also result in an increase in the flow of gases in the vicinity of the furnace wall tending to increase both damage to the body of the furnace by gas attack and detachment of material stuck to the furnace wall. This is likely to further aggravate fluctuations in the furnace thermal state.

It will be clear from the preceding paragraphs that we consider the way in which coke and ore are charged into the furnace, and especially the distribution in the thickness of the layers, to exercise a profound influence over the conditions indicative of movement of gas and charge within the furnace and of the conditions of the thermal state. It would therefore be of considerable value to be able to determine these factors with any certainty, both as an end in itself and also for control of the operation of the blast furnace.

However, it will be appreciated that a blast furnace constitutes a high temperature reaction furnace of substantial size covered with a thick layer of refractory material so that it has heretofore been quite difficult to form any very certain idea of the way in which the packed layers are disposed within the furnace. Such means as have been available heretofor in an effort to establish behaviour within the furnace are relatively crude. Thus, in one arrangement, the head of a chain or a wire was caused to descend to the surface of the charge immediately following charging thereof to measure the depth of the charge from a standard line so that some idea of the distribution of thickness could be obtained by taking the values so measured together with the similar values obtained after charging subsequent layers. Additionally, microwave detection systems have been suggested for measuring the distribution and shape of the surface of charges from the top of the furnace. However, a series of model experiments have shown us that the distribution and shape of the charges below the charging level are subject to change to a substantial degree due largely to disparity in the physical and other properties or ore and coke, including the difference in the momentum of ore and coke at the time of charging.

We believe that the disparity between distribution and shape of layers below the charging level from that established at charging may result from a number of factors. The angle of repose of coke is larger than that of ore and we find that the angle of inclination of the surface of a layer of coke is greater than for that of a layer of ore immediately following charging. However, when a charge of coke is followed by a charge of ore, the momentum of the ore is larger than that of the coke by as much as 3 to 4 times and the impact tends to push the layer of coke previously charged towards the centre of the furnace or towards the furnace wall depending on the previous angle of inclination. As a result, the surface of the layer of coke becomes rather more flat. Thus, the shape and configuration of the coke layer immediately following charging does not represent a good guide to the shape and configuration of the same layer subsequently. A further factor affecting the angle of inclination of the coke is that of bulk specific gravity, that of coke being approximately 0.5 while that of ore being approximately 2 so that the coke tends to be pushed upwards by the lifting power of gases rising from below during operation of the furnace.

Another attempt which has been made to gain knowledge about the behaviour of the charges within the furnace has been to fit a magnetic sensor to the furnace wall to detect the behaviour of the charges in the vicinity of the wall. This arrangement has proved of limited value since it can tell little about the behaviour of the charge except within a few score centimetres of the furnace wall. It has heretofore generally been regarded as quite beyond practicability to gain any direct knowledge of the behaviour of the layers of charge within the body of a conventional blast furnace with a diameter of 10 metres or more.

We should also mention other attempts at indirectly sensing the behaviour of the charges, as by measuring the distribution of the temperature or the distribution of the composition of gases at the top of the furnace with a horizontal probe or as by measuring the pattern of temperature on the surface of the charge from the top of the furnace by employing a,n infra-red camera and then deducing the distribution of the flow of gases or the distributlon' of the charges in the furnace from these results in an indirect manner. We have found that the adoption of such methods may provide erroneous information. In following the gas temperature distribution method, when the temperature is high or when the CO-to-CO2 ratio in the gases is beyond a predetermined level, the layer of coke is judged to be thick and the layer of ore to be relatively thin in the layer below the measuring point, and accordingly it is deduced that flow velocity of the gases in the said region is sufficiently high and permeability relatively favourable. However, when the temperature detected is law even when the gas flow velocity is sufficiently high, the detected temperature of the gases at the top of the furnace may be rather lower than the actual level. Furthermore, when the thickness of the layer of ore is relatively large, the CO-to-CO2 ratio is prone to be taken as being beyond a reasonable level when the velocity of downward movement of the charges in the measurement region is slow or when the temperature is so low that indirect reduction of ore by the CO gas is checked.

In accordance with a first aspect of the present invention, we provide a method of monitoring a blast furnace during operation thereof by inserting one or more magnetic sensors into one or more hollow tubes running through the body of the furnace at positions below the charging level for raw material to the furnace, and detecting variations in electrical outputs from said magnetic sensor(s) spatially and/or with time.

Suitably a plurality of the said sensors are disposed within the furnace and are coupled to signal processing means for processing the electrical outputs thereof.

In a second aspect of the present invention, we provide a method for controlling the operation of a blast furnace, in which a monitoring method as aforesaid is employed, the signal processing means being effective to process the electrical output signals from a plurality of the said sensors spatially distributed within the furnace below the charging level for raw material thereto to provide a map or distribution of one or more of velocity of downward movement of the charge, thickness of layers of the charge and angle of inclination at the boundary between layers of the charge at different locations within the furnace, means being provided for automatic control of one or more of the operations of charging fresh material into the furnace, blowing air into the furnace and controlling pressure at the top of the furnace on the basis of differences or divergence between the determined map or distribution and a predetermined standard or model map or distribution.

In a further and alternative aspect of the present invention, there is provided a blast furnace provided with one or more hollow tubes running through the body there below the charging level for raw material to the furnace, and one or more magnetic sensors position ble at one or more measurement locations in said hollow tube(s) and arranged to provide electrical output signals indicative of the local magnetic environment at said measurement location(s).

The invention is hereinafter more particularly described by way of example only with reference to the accompanying drawings, in which: Figures la and ib are schematic longitudinal sectional views through blast furnaces in accordance with the present invention Figures 2a, 2b and 2c are diagrammatic illustrations of the passage of layers of ore and coke past a particular magnetic sensor; Figures 3 and 4 are partial longitudinal sectional views through a blast furnace illustrating diagrammatically two different means by which magnetic sensors are moveable in the hollow tubes; Figures 5 and 6 are horizontal sectional views schematically illustrating the positioning of hollow tubes through the furnace; Figures 7, 8a and 8b are cross-sectional views showing the internal construction of three different embodiments of hollow tube; Figure 9 is a scrap longitudinal sectional view schematically illustrating the manner in which a hollow tube arrangement as shown in Figure 8b is fitted through the wall of the furnace Figure 10 is a graph setting out typical results from four magnetic sensors in the course of a method in accordance with the present invention; Figure 11 is a schematic part longitudinal sectional view showing the positions of the magnetic sensors giving the results in Figure 10; Figure 12 is a graph showing on an enlarged scale part of the waveform output from a magnetic sensor for the purpose of explanation; Figure 13 is a graph showing several magnetic sensor outputs for the purpose of explaining how velocity of downward movement, thickness of the layers and the angle of inclination of the charges may be calculated; Figure 14 is a further schematic partial longitudinal sectiqnal view of a blast furnace illustrating how sensors may be positioned to obtain outputs such as shown in Figure 13; Figure 15 is a diagrammatic representation showing how the results obtained from the magnetic sensors may be used to plot the thickness of the layers of charge and the configurations taken up by these layers in the furnace; and Figure 16 is a similar diagrammatic representaion plotted from different results obtained under different circumstances.

Within the body 1 of the blast furnace schematically illustrated in Figure la, will be seen alternating packed layers 2 and 2 respectively of iron ore and coke. The iron ore layers 2 (which may include sintered ore) and the coke layers 3 will move downwardly within the furnace as its operation proceeds. A hollow tube 4 is shown running through the body of the furnace at a position below the charging level 5 for raw material to the furnace and at least one magnetic sensor 6 is positioned in the interior of the hollow tube 4. The changing magnetic environment as the alternating layers of ore and coke pass the position of a sensor 6 will be detectable as magnetic field changes at the sensors producing an output signal which passes to a signal processing means 7 in which the raw signals may be processed in conventional fashion by amplification. matching of waveform and the like. As will be explained in more detail below the processing means may be arranged from the results of several sensors to provide a determination for the velocity of downward movement of the charges. the thickness of the layers and the configurations they take up in the body of the furnace.

Figure ib shows another embodiment of furnace constructed in accordance with the present invention. As before. alternate layers of ore 2 and coke 3 are charged at the top of the furnace. Air is blown into the bottom of the furnace through a tuyere 9 and the coke is completely consumed in the combustion zone 10 immediately in front of the tuyere 9. The layers of charges thus move downward. As the upper surface layer 5 of the charge reaches a predetermined level. charging of the next layer of material is effected. Thus. the uppermost level 5 is maintained reasonablv constant. A plurality of the hollow tubes 4 run through the body of the furnace at different horizontal levels below the charging level 5 and may be parallel. as shown. Each said hollow tube 4 has a selected number of magnetic sensors 6 mounted in the interior thereof at selected sensing positions. Suitably the sensors are arranged so that a sensor in the lower tube is immediately below a corresponding sensor in the upper tube. It is desirable that the hollow tubes 4 pass through the body of the furnace at a position where the temperature of the charge will be below the Curie point of the ore.

The separation between individual hollow tubes 4 in the vertical direction is desirable less than the thickness of a layer of coke or of a layer of ore. and most suitable within the range of 150 to 300 mm. Preferably the hollow tubes 4 run diametrically through the central or axial region of the furnace. As before, each magnetic sensor will detect fluctuations in the magnetic flux density of the local magnetic field dependent on the downward movement of the charges. The magnetic sensors are connected to a processing means 7 for processing the electrical outputs thereof. Indicator means 8 may be connected to the processing means 7 for the purpose of displaying or recording the signal output resulting from the processing of the fluctuating magnetic signals.

We shall now explain how the magnetic fluctuations arise and may be detected.

Figures 2a, 2b and 2c show a magnetic sensor 6 having an exciting section 6a for creating a localised magnetic field which will magnetize any ferro-magnetic material in the vicinity, and also a magnetic detecting section 6b which is effective to detect changes in the vector components of the resultant magnetic field. The detecting section 6b is positioned so as to be intersected at right angles by the longitudinal axis X of the exciting section 6a and also so as to be parallel to the downward direction of movement of the charge. Figure 2a shows what happens when the centre of the detecting section 6b is positioned midway between the upper and lower boundary layers of a coke layer 3. The resulting magnetic field as detected by detecting section 6b shows axial symmetry about the longitudinal axis X. In effect, any disturbance to the exciting magnetic field created by section 6a caused by the next adjacent preceding and next adjacent following ore layers 2 are equal and opposite and cancel each other out. As the layers of coke and ore proceed on down, the situation shown in Figure 2b will arise when a boundary layer is positioned halfway down the detecting section 6b.

Because of the magnetic permeability of the ore 2, the magnetic field will be substantially deflected into the ore layer. The upward and downward components of the magnetic field detected at section 6b will no longer be equal; section 6b will detect an upward deflecting magnetic field and produce a positive value signal. As the charge layers move further downward, the central region of an iron ore layer 2 will confront detecting section 6b and the resultant magnetic field will again be symmetrical about the longitudinal axis X. The output from the magnetic sensor 6 will again become zero. As the charged layers continue to descend the situation depicted in Figure 2c will arise, being the converse of Figure 2b.

The detector section 6b will detect a downward deflecting magnetic field and the magnetic sensors 6 will provide a negative value output signal as an input to the processing means 7.

In the arrangement described hereinabove, each magnetic sensor 6 comprises an exciting section 6a which creates an exciting magnetic field in an active manner in order to detect fluctuations in the local magnetic environment. However, as is well known, some iron ore is itself magnetised having a fairly high level of magnetising force. In such a case, movement of the alternating layers of coke and ore may be detected in a manner generally similar to that described above, by detection of the unmodified magnetic field of the iron ore itself.

The magnetic sensors 6 used in the practice of the present invention should be especially adapted for detecting fluctuations in the vector components of the magnetic force created by an exciting section of the sensor or the magnetic field borne by the iron ore itself. A magnetism-to-electricity transducer, a gaussmeter, or a magnetism detector may be employed for the sensor 6. Especially effective magnetic sensors are those of the SMD (Sony Magneto Diode) type, those of the Hall element type making use of the Hall effect, those of the search coil type, those of the D.C.-A.C flux-gate type, and those of the electric resistance effect type. To obtain an output featuring stability and high sensitivity, we prefer to employ a magnetic sensor of the magnetic multivibrator type.

In the sensor 6 referred to in connection with Figures 2a to 2c, the detector section 6b may be referred to as a magnetometer. In actual fact, it is provided with a magnetic sensitive section (i.e. a magnetometer in the narrow sense) and a circuit driver section providing a corresponding oscillating electrical signal which passes to the processing means 7. The exciting section 6a will usually comprise a permanent magnet and may be of the kind in which a coil is wound round a magnetic core and is supplied with A.C. or D.C. power.

Alternatively, an exciting coil alone without a magnetic core may be employed. The selection of an appropriate exciting section 6a may be made by those with skill in this art on the basis of such criteria as the intensity of the exciting magnetic field required, the dimensions of the magnet, and whether or not the arrangement selected is easy to handle.

Turning now to Figure 12 which shows on an enlarged scale part of the waveform output from one magnetic sensor, at the time t, the upper half of the detector section 6b of a sensor 6 is confronted by a coke layer 3 while the lower half of the detector section 6b is confronted by a layer of ore 2, the boundary between the layer of ore and the layer of coke coinciding with the position of the sensor. As the ore 2 has a substantial magnetic permeability, the lines of magnetic force will be substantially deflected in the direction of the ore (i.e. in the downward direction) and there will be a distinct difference between the vector components of the magnetic force in the upper and lower parts of the detector section 6b. In consequence, the magnetic sensor provides a negative value signal 21. When the centre of the coke layer 3 passes the central axis X of the detector section 6b (as described hereinabove with reference to Figures 2a to 2c) as the layers descend in the furnace, the lines of magnetic force become symmetrical relative to the axis X. Consequently, a zero signal 22 is provided at the output of the magnetic sensor. With further descent of the layers of charge, the next boundary between coke 3 and ore 2 will arrive at the centre of the detector section 6b so that the upper half thereof is exposed to ore and the lower half to coke at a time t2. The lines of magnetic force are now substantially deflected towards the upper ore layer 2 and a positive value signal 23 results from the output of the magnetic sensor 6. With further downward movement of the charge the centre of the ore layer 2 will reach the detector section 6b and at that time a zero signal 24 will be provided at the output from sensor 6. To sum up, the signal output from the magnetic sensor 6 passes through maxima and minima at the boundaries between the layers of ore 2 and coke 3 at respective times t1, t2, t3 Alternative arrangements are possible in which the signal output from the sensor 6 is reduced to zero at the boundary between respective layers.

As mentioned above in relation to Figure ib, when a plurality of hollow tubes 4 are employed, we prefer the sensors to be arranged in the the two tubes so that a sensor in each tube below the uppermost is immediately below a corresponding sensor in the uppermost tube. The position of the sensors may be adjustable in either of the manners shown in Figures 3 or 4. In Figure 3 a sensor 6a is shown connected to a wire or chain 9 the ends of which are mounted on respective reels or drums 10 for winding on the wire or chain in the appropriate direction of travel of the sensor 6a either by means of driving gear or manually.

In the Figure 4 arrangement a sensor 6b is mounted at the end of a rod 11 arranged for travel in the forward and rearward directions by means of a pinion gear 12 adapted to be driven. Other arrangements are, of course, possible for moving the rod 11, including cylinder drive systems or screw drive systems. Other arrangements for moving magnetic sensors in the tubes will readily occur to those skilled in the art.

As has been described above with reference particularly to Figures la and lb, in the practice of this invention, one or more hollow tubes 4 run through the body of the furnace at a position below the charging level 5 for raw material to the furnace. The arrangements of Figures la and 1b are schematic. In practice, hollow tubes 4a may run through the body of the furnace so as to cross each other at an angle, such as a right angle, as shown in Figure 5. Again, it is not necessary for the tubes to extend diametrically across the furnace. Tubes 4b may be mounted in parallel as shown in Figure 6. The tubes may be mounted in a succession of vertically separated stages. The number and arrangement of the hollow tubes is a matter for selection depending on various criteria, including the size and configuration of the furnace, the number and placing of the magnetic sensor(s) is also a matter to be decided depending upon the results required and the particular arrangement chosen for the hollow tubes.

Figure 7 shows a sectional view through one embodiment of suitable hollow tube 4. Tube 4c is hollow and cylindrical, the magnetic sensor 6 is fitted in place in its interior. A protective cover 13 is provided for preventing or mitigating wear on the tube and to assist in smooth downward movement of raw material past the tube. The hollow tube 4c may be made of such materials as stainless steel, copper, or other non-magnetic substances. We prefer the use of a stainless steel pipe. The configuration of the tube is not limited to circular cylindrical pipes. Other shapes, such as triangular cylinders or quadrangular cylinders may be employed. The protective cover 13, though highly preferred, is not absolutely essential. The hollow tube 4 should, however, be made of a material sufficiently durable to withstand the downward movement of the charge therepast and the superposed load of charge.

The lower region of the furnace tends to become hotter than the uppermost reg ambient temperature and the shape of the tube.

The Figure 8b arrangement shows a composite hollow tube 4d allowing magnetic sensors to be positioned in different horizontal planes through the body of the furnace. Respective sensors 6 are shown positioned within inner hollow tubes 4d1, 4d2 in upper and lower stages.

Each tube 4dl, 4d2 has a corresponding circumlocated outer hollow tube 4d3, 4d4. Tubes 4d3, 4d4 are connected together in appropriate fashion by fixing ribs 14. As will be seen from the drawing, the lower tube 4d4 is somewhat larger than the upper tube 4d3 and has a further tube 4d5 located therewithin. Tube 4d2 is located in the space intermediate tube 4d5 and 4d4. A passage 15a for a cooling agent is defined between tubes 4d3 and 4d1. A further passage 15b for cooling agent is defined between tubes 4d5 and 4d4 and surrounding tube 4d2. Suitable supporting plates 16a, 16b and 16c are provided for supporting tubes 4d1, 4d2 and 4d5 within their respective outer tubes. The use of tube 4d5 means that the total volume of cooling fluid required for passage 15b is reduced. We find that this reduction does not reduce the effectiveness of the cooling fluid. The passageway defined within tube 4d5 may be used to measure the temperature of the furnace and to sample furnace gases. Figure 9 shows how the composite tube of Figure 8b may pass out through the wall of the furnace.

It will be appreciated that in both the arrangements of Figure 8a and Figure 8b the magnetic sensor is cooled indirectly. If the nature of the sensor is such that it will not be affected by the cooling agent, the arrangement of Figure 7 may be employed with cooling agent, passed through the space 4cl.

It will further be appreciated that the arrangements described with reference to Figure 7, Figure 8a and Figure 8b are examples only. Those with the necessary skills will readily be able to devise alternative arrangements which allow for the sensors and/or hollow tubes to be cooled in situ within the body of the furnace.

Turning now to Figure 10, there is shown a graph setting out typical results from four magnetic sensors 6s1...6s4 mounted with equidistant spacings of 890mm. within a hollow tube 4 arranged 4100mm. below the stock line S.L. of a blast furnace having an internal volume of 2800 m3, as schematically indicated in Figure 11. The "stock line" S.L. lies at a position 1 m below the lower end of the lower bell of the furnace. In the graph of Figure 10, the ordinate indicates elapsed time, each scale interval representing a span of 12 minutes; and the abscissa indicates the level and direction of the output signals from the sensors. For each sensor the right hand direction represents a positive value (+) and the left hand direction represents a negative value (-). The maxima and minima in the respective output signals occur with the passage of a boundary layer between a respective iron ore layer 2 and a coke layer 3. as discussed hereinabove.

We shall now explain in more detail how the sensor outputs may be employed to derive the velocity of downward movement of the charge, the thickness of layers of charge and the configuration of the layers of charge within the furnace.

Figure 13 shows a segment of the signal output fll from a magentic sensor 611 positioned vertically beneath an upper magnetic sensor 6ul and having a signal output fu. The signal output fu. of a second magnetic sensor 6u2 positioned at the same vertical height in the furnace as sensor 6u, is also shown. Figure 14 shows the arrangement of upper sensors 6cult 6u2. 6up,* 6u4.. and corresponding lower sensors 61l. 612, 613, 614... in the furnace. In Figure 13 the ordinate represents elapsed time while the abscissa is the magnitude and direction of the sensor output. positive values being shown to the right and negative values to the left of the sensor output centre line in each case. As discussed hereinabove the maxima and minima in the signal outputs represent the boundaries between layers of iron ore 2 and coke 3. The passage of the same boundary layer past adjacent sensor positions either in the same horizontal plane or one vertically beneath the other may thus be readily followed. For respective upper and lower corresponding sensors 6ui and 61i (as for example the sensors 6u1 and 61 shown in Figure 14), the velocitv Vi of downward movement of the charge may be calculated by application of the formula set out below if the mean value Ti of the time differences Ti. j (j = 1.2.3. .) between corresponding peaks (i.e. maxima and minima) of the signal output curves fui and fli of the respective upper and lower magnetic sensors 6ui and 6li is calculated. The mean value Ti may be calculated over a suitable predetermined period of for example 30 minutes or one hour. The formula to be applied is Vi = H/Ti (1) Here. Vi Velocity of descent of the charge along a designated vertical line H Distance between upper and lower magnetic sensors along said line.

As an alternative to the above described approach to the calculation of the velocity of downward movement of the charge in which the respective time differences Ti,j, may be summed sequentially, the formula (1) set out above may be applied with a time difference value Ti which is selected as that value of ti which maximizes the mutual correlation coefficient g (Ti) defined by the formula (2) set out below.

r g (ti) = T fui (t). fli (t-li) dt (2) fO Here, g (tri) : Mutual correlation coefficient fui, fli : Respective signal outputs of the upper and lower magnetic sensors t Time T : Correlation operation time interval (set time) T : Time difference between peaks.

In this way calculated values V1, V2, V3... for the velocity of downward movement of the charge may be derived for a series of sensor positions (i = 1,2,3...) to provide a map or distribution of velocity of downward movement of the charge in the furnace.

There may additionally be provided in the blast furnace one or more charging level meters such as sounding level meters, microwave level meters, or ultrasonic level meters, these charging level meters being positioned on the body of the blast furnace to provide additional calculated values for the velocity of downward movement of the charge within the furnace by directly detecting downward movement of the top surface of the packed layers of the charge within a predetermined period.

As we have explained above the boundary between an iron ore layer 2 and the adjacent coke layer 3 can be readily ascertained from the outputs of the respective magnetic sensors, since their output signals pass through a maximum or a minimum at each boundary. Thus, the time of passage Lto of a layer of ore past a particular magnetic sensor can readily be ascertained. Equally well. the time of passage Atc of a layer of coke past the same magnetic sensor can be readily established. From the calculated velocity of downward movement of the charge Vi at a particular location i, which velocity is calculated according to one of the methods described above, the thickness ho of a particular layer of ore and the thickness hc of a particular layer of coke may be calculated by the following formulae (3) and (4) ho = Vi. Ato . . (3) hc = Vi. Atc . . (4) Calculations according to the formulae (3) and (4) may be performed for individual layers of iron ore 2 and coke 3. Alternatively, a mean thickness may be calculated over a set period of time by calculating the mean time for a plurality of layers of the respective materials to pass the measuring point i.

In any event. the procedure described may be used to establish a map or distribution of the thickness of the layers of ore and coke at different measuring positions ins the furnace.

The use of magnetic sensors as described hereinabove also enables a determination to be made of the inclination of the boundary surface between two layers of charge in the neighbourhood of two adjacent magnetic sensors spaced horizontally from each other. For sensors at positions i. i+1 the angle of inclination 0i, i+1 is given by the formula (5): Oi. i+1 = tan-1 (ai i+1/Li. i=1). (5) Li. i+1 is the distance horizontally between the sensors at positions i. i+1 and cti, i+1 is the deflection downwards of a particular boundary surface at position i+1 relative to position i.

This deflection mav itself be calculated according to the following formula (6): ai. i+1 = Vi+1. ATi + ssi. its.. ..(6) Here. Vi+1 Velocity of the downward movement of the charge at the point i+1 ATi Mean time between corresponding peaks and troughs in the respective signal outputs fui and fui+1 of points i, i+1.

Pi,i+l : Additional slippage or deflection of the boundary surface in the downward direction at the point i+1 relative to point i.

The velocity of downward movement of the charge at point i+ 1 may be calculated according to any of the methods described above. The mean time ATi between corresponding peaks and troughs (maxima and minima) in the respective signal outputs fui and fui+1 may be calculated by various methods. These include the establishment of a mutual correlation coefficient g (nazi) for the signal outputs fui and fui+ 1 in much the same way as described above for the determination of the time difference Ti between corresponding maxima and minima of the signal outputs fui and fli of respective upper and lower magnetic sensors located along the same vertical line. Alternatively, the method of double charging may be employed. In this method the weight of charge is temporarily increased by doubling it. In this way a particular boundary surface BS may readily be picked out from the output signals fui and fui+ 1 and the time difference may be directly determined. This process may be repeated a number of time to establish a mean value ATi.

The above described calculations may be repeated for various values of i to produce a distribution or map of the inclinations between the boundary layers within the furnace.

We have now explained how methods in accordance with the present invention may be employed to establish a map or distribution of one or more of velocity of downward movement of the charge, thickness of layers of the charge and angle of inclination of the boundary between layers of the charge at different locations within the furnace. The results may be displayed on an indicator. Again, means may be provided for automatic control of one or more of the operations of charging fresh material into the furnace, blowing air into the furnace and controlling pressure at the top of the furnace on the basis of differences or divergence between the determined map or distribution and a predetermined standard or model map or distribution.

We shall now describe some practical arrangements and the results achieved with reference to Figures 14 et seq. Figure 14 shows magnetic sensors 6u1 to 6u4 and 611 to 614 mounted in place in two stages separated vertically, the sensors in each horizontal level being separated from each other by spacings of the order of 890 mm. The tube 4 passes diametrically across the furnace at a position 4100 mm below the stock line S.L. which is itself located 1 m below the lowest portion of the large bell 18. The internal volume of the particular blast furnace was 2800m.The procedure outlined above for producing a map or distribution of the velocity of downward movement of the charge, the thickness of the layers of iron ore 2 and coke 3 and the configuration of those layers (angle of inclination of the respective boundary surfaces) was followed for the magnetic sensors 6u1 to 6u4 and 61l to 614. The results are shown in Figure 15. On this graph, the ordinate represents the height and thickness of individual layers and the velocity of downward movement, taking the lower surface of the standard coke layer 3 in the vicinity of the wall of the furnace as a basis for evaluation. The abscissa represents the radius of the furnace and shows the respective radial locations 1,2,3,4 of the upper and lower magnetic sensors 6u1 to 6u4 and 61l to 614.

In the example of Figure 15 each charge of iron ore comprised 69.4 metric tonnes and each charge of coke was of 18.5 metric tonnes. As will be clear from Figure 14, a moveable armor 17 is disposed in the path of charge falling from the large bell 18. The orientation of the armor 17 is adjustable by means of a series of notches. As the angle at which charge strikes the armor is adjusted by means of the notches, so is the falling position of the charge.

The pattern of notch settings for the moveable armor during the charging of coke C and charging of ore 0 is indicated in Figure 15, from which it will be seen that the notch settings for the charging of coke C was twice set at 5.5 notches and once set at five notches while the notch setting during the charging of ore 0 was three notches throughout. Different kinds of coke were selected for five notch charging and 5.5 notch charging. As is clear from Figure 15 the larger the number of the notch setting, the nearer to the centre of the furnace will the charge fall. Thus, under the working conditions adopted for Figure 15, coke is charged nearer to the centre of the furnace than iron ore. In this case the layer of iron ore was deliberately made relatively thicker in the vicinity of the furnace wall while the layer of coke was deliberately made relatively thicker at the centre of the furnace. Hollow tubes of the kind shown in Figure 8b were used. The experimental results are depicted in Figure 15 in which the layers of iron ore 2 are shown hatched.

The velocity distribution shown is for the coke layer charged with five notches. As is apparent from the drawing, the thickness and- angle of inclination varies substantially.

There is no linear progression; there is a sharp change in inclination at an intermediate position apart slightly from the wall of the furnace, the inclination being relatively gentle close to the furnace wall. It will be seen that the velocity of downward movement is substantially higher at the centre of the furnace that it is closer to the wall of the furnace.

As will be apparent, the production of a graphical representation such as that shown in Figure 15 enables any departure from the desired or standard distribution of velocity of downward movement of the charge, thickness of layers of the charge and angle of inclination at the boundary between layers of the charge at different locations within the furnace to be readily established. The operation of the blast furnace may be varied in accordance therewith, by, for example, manually operated control of one or more of the operations of charging fresh material into the furnace, blowing air into the furnace and controlling pressure at the top of the furnace. Alternatively, and as preferred, automatic control rather than manually operated control may be provided for these functions. A fundamental improvement in the conditions of operation of the furnace can be achieved by control of the conditions of charging, including the depth of charging. The furnace thermal state, and conditions indicative of the movement of charge (permeability of the furnace) may be adjusted in accordance with the divergence from the standard distributions as aforesaid by control of the blast conditions, including an increase/decrease in the blast volume, modification of the flow rate of heavy oil, the flow rate of oxygen, the blast temperature, or the blast humidity. The velocity of flow of gas in the furnace may be adjusted for best results and/or an improvement in the reduction efficiency by the gases can be achieved.

We shall now explain something more of the theory behind the controlling adjustments which may be made to the furnace.

Control of charging may be roughly classified into two categories, namely control of the amount of charge and control of the distribution of charge. The control of the amount of charge is simple quantitative control in terms both of iron ore 2 and coke 3. The quantity of iron ore is reduced or the quantity of coke increased when the mean velocity of downward movement of the charge is in excess of the preset standard. To put it another way, the ore-to-coke ratio is reduced to a lower level, thereby stabilizing the furnace thermal state in a preset condition, since if the velocity of downward movement of the charges should increase despite the blast conditions, including the blast volume at the tuyere being constant, heat exchange between the charge and gases rising within the furnace and the reduction of the iron ore by carbon monoxide and Hydrogen will fall short resulting in a lowering of the thermal state of the furnace.

A second objective of the control of the charge is to improve its distribution. When a blast furnace is not provided with a charge distribution control device such as a moveable armor, such as that shown as 17 in Figure 14, charge control plays an important role as a charge distribution control means. Since, as mentioned above, the angle of repose of coke is generally smaller than the angle of repose of iron ore, a change in the quantity of coke charged each time through the top of the furnace (hereinafter referred to as "the coke base") results in a change in the distribution of the thickness of the layers of charge diametrically across the furnace even if the ore-to-coke ratio remains the same. The smaller angle of repose of coke makes it tend to flow toward the centre of the furnace while ore with a rather larger angle of repose is prone to be deposited near the furnace wall. When the coke base is small, the quantity of coke and the quantity of iron ore to be charged on each occasion is small and mainly coke will be charged into the centre of the furnace and mainly ore near the furnace wall. This improves the permeability in the centre of the furnace and reduces the flow of gases adjacent the furnace wall. When the coke base is enlarged, the quantity of coke deposited adjacent the furnace wall increases though the relative amount remains the same and the quantity of ore caused by the coke to flow towards the centre of the furnace increases which tends to make the distribution of thickness of layers more uniform. Thus, the distribution of the thickness of layers of the charge in the radial direction of the furnace detected by a method in accordance with the present invention may be used to control the coke base. Alternatively, the above described moveable armor may conduct a direct control of the position of drop of ore and coke from the top of the furnace and thus provides a useful means for effecting control of distribution of the thickness of the layers of charge.

Now, as to the blast conditions: if the velocity of downward movement of the charge should exceed the preset standard value, it may be reduced (or to put it otherwise, proper control of the blow may be exercised) either by reducing the blast volume or the flow rate of oxygen, or by increasing the flow rate of heavy oil to stabilize the thermal state. Of course, much the same effect may be achieved by other means of blast control such as control of the blast temperature and control of the blast humidity. If the distribution of the thickness of the layers of charge or their configurations is lacking in uniformity in the radial direction of the furnace. the blast volume and/or the flow rate of oxygen may be modified to render the resistance to gas permeability non-uniform in the radial direction of the furnace. If the quantity of coke present at the centre of the furnace is large and the flow of gases in the furnace is concentrated in the centre of the furnace, the velocity of gas flow may be lowered by either reducing the blast volume or by elevating the top gas pressure whereby the distribution of the flow of gases in the radial direction of the furnace can be made more uniform. Lastly, with regard to the control of top gas pressure, the situation is broadly analogous to the control of the blast. If the distribution of the thickness of the layers of charge or the configuration of the layers is not uniform in the radial direction of the furnace, or if a considerable difference exists between the preset standard value of the top gas pressure and that present, increase of that pressure tends to render the distribution of the velocity of gas flow in the radial direction more uniform, whereby reduction efficiency in the furnace is increased, permeability may be improved and the rate of fuel consumption may be reduced. On the other hand, if the flow of gases in the central region of the furnace is to be accelerated, top gas pressure may be reduced or the blast volume increased.

Figure 16 is a diagrammatic representation generally similar to Figure 15 showing the distribution of the thickness of the layers of ore 2 and coke 3 and the general configuration of such layers for a furnace charged under somewhat different conditions. The positions of four magnetic sensors 6u1 ... 6u4 are indicated relative to the centre of the furnace and the furnace wall, the spacings employed being 890mm within a hollow tube 4, the hollow tube being 4,100mm below the stock line, the furnace having an inner volume of 2800m3. The illustrated results were employed with a charge of 72.4 metric tonnes of ore in each charge of ore and 18.5 metric tonnes of coke in each charge of coke. The pattern of notch settings for the moveable armor is indicated in Figure 16. Alternate 5 notch and 4.5 notch settings were employed for the charging of coke while the charging of ore was conducted with a 3 notch setting throughout. The results showed the higher the notch setting the nearer the centre of the furnace the charges fall. Thus, the coke 3 tended to be charged nearer to the centre of the furnace than the ore 2, the coke layers were relatively thicker at the centre of the furnace, and the ore layers were generally thicker adjacent the wall of the furnace. In this case, a single stage hollow tube 4 was employed and the velocity of downward movement of the charges was taken as 5,930 mm per hour calculated by means of a sounding level meter.

In Figure 16, the hatching indicates layers of ore 2, the blank portions representing layers of coke 3. In this case, the standard surface from which thicknesses and inclinations are measured is taken as the upper boundary surface of the 5 notch setting coke. The drawing indicates the averaged value of the thicknesses and inclinations of the layers relative to the standard surface, the averaging being made over a period of approximately 4 hours.

It will be seen that as little a change in the notch setting of the moveable armor as 0.5 notch results in a considerable change in the distribution of the thickness of the layers and the angle of inclination. It will also be seen from Figure 16 that the inclination of the layers is relatively steep at an intermediate radial position in the furnace, being relatively gentle in the vicinity of the furnace wall and also at the centre of the furnace.

WHAT WE CLAIM IS: 1. A method of monitoring a blast furnace during operation thereof by inserting one or more magnetic sensors into one or more hollow tubes running through the body of the furnace at positions below the charging level for raw material to the furnace, and detecting variations in electrical outputs from said magnetic sensor(s) spatially and/or with time.

2. A method according to Claim 1, wherein the one or more magentic sensors are moveable in the said one or more hollow tubes.

3. A method according to Claim 1 or 2, wherein said one or more magnetic sensors are coupled to signal processing means for processing the electrical outputs thereof.

4. A method according to any preceding Claim, wherein the or each said hollow tube is provided with a cooling agent circulation system.

5. A method according to any preceding claim, wherein a plurality of the said hollow tubes run through the body of the furnace at different horizontal levels below the charging level, whereby to detect the passage downwardly of a boundary layer between layers of charged material having different magnetic properties.

6. A method according to any preceding claim, wherein a plurality of the said hollow tubes extend through the body of the furnace at angles to each other.

7. A method according to any preceding claim, wherein said one or more hollow tubes include at least two tubes extending parallel to each other in the same vertical plane so that for each magnetic sensor located in the uppermost of the two tubes, there is present a magnetic sensor in the lower of the two tubes vertically beneath the first mentioned magnetic sensor, whereby the downward movement of layers of charged material having different magnetic properties may be monitored.

8. A method according to any preceding claim, wherein the or each said hollow tube is double walled, comprising a first tube located within a second tube, the magnetic sensor(s) being located within the first tube(s).

**WARNING** end of DESC field may overlap start of CLMS **.

Claims

**WARNING** start of CLMS field may overlap end of DESC **. furnace is concentrated in the centre of the furnace, the velocity of gas flow may be lowered by either reducing the blast volume or by elevating the top gas pressure whereby the distribution of the flow of gases in the radial direction of the furnace can be made more uniform. Lastly, with regard to the control of top gas pressure, the situation is broadly analogous to the control of the blast. If the distribution of the thickness of the layers of charge or the configuration of the layers is not uniform in the radial direction of the furnace, or if a considerable difference exists between the preset standard value of the top gas pressure and that present, increase of that pressure tends to render the distribution of the velocity of gas flow in the radial direction more uniform, whereby reduction efficiency in the furnace is increased, permeability may be improved and the rate of fuel consumption may be reduced. On the other hand, if the flow of gases in the central region of the furnace is to be accelerated, top gas pressure may be reduced or the blast volume increased. Figure 16 is a diagrammatic representation generally similar to Figure 15 showing the distribution of the thickness of the layers of ore 2 and coke 3 and the general configuration of such layers for a furnace charged under somewhat different conditions. The positions of four magnetic sensors 6u1 ... 6u4 are indicated relative to the centre of the furnace and the furnace wall, the spacings employed being 890mm within a hollow tube 4, the hollow tube being 4,100mm below the stock line, the furnace having an inner volume of 2800m3. The illustrated results were employed with a charge of 72.4 metric tonnes of ore in each charge of ore and 18.5 metric tonnes of coke in each charge of coke. The pattern of notch settings for the moveable armor is indicated in Figure 16. Alternate 5 notch and 4.5 notch settings were employed for the charging of coke while the charging of ore was conducted with a 3 notch setting throughout. The results showed the higher the notch setting the nearer the centre of the furnace the charges fall. Thus, the coke 3 tended to be charged nearer to the centre of the furnace than the ore 2, the coke layers were relatively thicker at the centre of the furnace, and the ore layers were generally thicker adjacent the wall of the furnace. In this case, a single stage hollow tube 4 was employed and the velocity of downward movement of the charges was taken as 5,930 mm per hour calculated by means of a sounding level meter. In Figure 16, the hatching indicates layers of ore 2, the blank portions representing layers of coke 3. In this case, the standard surface from which thicknesses and inclinations are measured is taken as the upper boundary surface of the 5 notch setting coke. The drawing indicates the averaged value of the thicknesses and inclinations of the layers relative to the standard surface, the averaging being made over a period of approximately 4 hours. It will be seen that as little a change in the notch setting of the moveable armor as 0.5 notch results in a considerable change in the distribution of the thickness of the layers and the angle of inclination. It will also be seen from Figure 16 that the inclination of the layers is relatively steep at an intermediate radial position in the furnace, being relatively gentle in the vicinity of the furnace wall and also at the centre of the furnace. WHAT WE CLAIM IS:

1. A method of monitoring a blast furnace during operation thereof by inserting one or more magnetic sensors into one or more hollow tubes running through the body of the furnace at positions below the charging level for raw material to the furnace, and detecting variations in electrical outputs from said magnetic sensor(s) spatially and/or with time.

9. A method according to Claim 7, wherein a comparison is made between the

fluctuating electrical output from the upper and lower magnetic sensors to determine the time difference between the occurrence of a peak or trough in the output signal from the upper sensor and the corresponding peak or trough in the output of the lower sensor; the velocity of downward movement of the charge along the vertical line between the upper and lower sensors is determined by taking the said time difference between corresponding peaks or troughs together with the distance between the upper and lower sensors; the time of passage of a particular layer of the charge past one or both of said upper and lower sensors is determined from the portion of said fluctuating output signal corresponding to the said layer of charge; and wherein the thickness of the said layer of charge is determined by taking the determined time of passage of the said layer past a said sensor together with the overall velocity of downward movement of the charge between the upper and lower sensors.

10. A method according to Claim 9, wherein the angle of inclination of the boundary between two layers of charges is determined by detecting the time difference between the passage of a peak or trough corresponding to the said boundary at one sensor and the corresponding peak or trough at another sensor spaced horizontally from the first, and taking the so determined difference together with the determined velocity of downward movement of the charge.

11. A method for controlling the operation of a blast furnace, the method comprising monitoring the said furance during operation thereof by a method according to Claim 3 or any claim appendant thereto, wherein the signal processig means is effective to process the electrical output signals from a plurality of the said sensors spatially distributed within the furnace below the charging level for raw material thereto to provide a map or distribution of one or more of velocity of downward movement of the charge, thickness of layers of the charge and angle of inclination at the boundary between layers of the charge at different locations within the furnace, means being provided for automatic control of one or more of the operations of charging fresh material into the furnace, blowing air into the furnace and controlling pressure at the top of the furnace on the basis of differences or divergence between the determind map or distribution and a predetermined standard or model map or distribution.

12. A method according to Claim 11, wherein the velocity of downward movement of the charge is calculated by application of the following formula: Vi = H/Ti... .. (1) (i = 1, 2, .

Here, Vi Velocity of descent of the charge along a designated vertical line H Distance between upper and lower magnetic sensors along said line Ti Mean value of difference in time Ti,j (j = 1, 2, 3, .) between corresponding peaks, during a predetermined period, in the respective signal outputs fui, fli of the said upper and lower magnetic sensors.

13. A method according to Claim 11, wherein that value of time difference Ti between corresponding peaks of the respective signal outputs fui, fli of upper and lower magnetic sensors located along the same vertical line within the furnace for which a mutual correlation coefficient g(Ti) determined according to the following formula (2) becomes a maximum is employed for calculating the downward velocity Vi of the charge along said line: g (Ti) = T fui (t). fli (t-Ti) dt (2) Here, g (li) Mutual correlation coefficient fui,fli Respective signal outputs of the upper and lower magnetic sensors t Time T Correlation operation time interval (set time) T Time difference between peaks.

14. A method according to Claim 11 or any claim appendant thereto, wherein the respective thicknesses of a layer of ore ho or of coke he in a predetermined location of the furnace are determined by the following formulae (3) and (4) respectively: ho = Vi.A to ..

he = Vi.A te ..

Here, ho Thickness of the layer of ore he Thickness of the layer of coke Vi Calculated velocity of downward movement of the charge at the location i Ato Time of passage of a layer of ore past a magnetic sensor Ate Time of passage of a layer of coke past a magnetic sensor.

15. A method according to Claim 11 or any claim appendant thereto, wherein the inclination of a boundary surface between two layers in the neighbourhood of two adjacent magnetic sensors spaced horizontally from each other is determined by calculating the mean time ATi between corresponding peaks or troughs in their respective signal outputs fui and fui + 1 and by the application of the following formula (5): 6i, i+1 = tan -1 (ai, i+1/Li, i+1) (5) Here, Oi, i+1 : Angle of inclination of the boundary surface in the region of points i,i+ 1 Li, i+1 : Distance horizontally between the magnetic sensors at points i,i+1 ai, i+1: Deflection downwards of the boundary surface at position i+1 relative to position i, which deflection is calculated according to the folowing formula (6): ai, i+1 = Vi+1, ATi + i,i+1 (6) Here, Vi+1 : Velocity of the downward movement of the charge at the point i+1 ATi : Mean time between corresponding peaks and troughs in the respective signal outputs fui and fui+1 of points i, i+1.

i, i+1 Additional slippage or deflection of the boundary surface in the downward direction at the point i+1 relative to point i.

16. A method according to Claim 15, wherein a particular boundary surface is identified by charging a substantially greater quantity of charge (either ore or coke) on one particular charge and monitoring the output signals to detect passage of the boundary layer associated with said substantially greater charge, the boundary layer inclination calculations being performed in respect of the said boundary surface and neighbouring preceding and following boundary surfaces.

17. A method of monitoring a blast furnace substantially as hereinbefore described with reference to the accompanying drawings.

18. A blast furnace provided with one or more hollow tubes running through the body thereof below the charging level for raw material to the furnace, and one or more magnetic sensors positionable at one or more measurement locations in said hollow tube(s) and arranged to provide electrical output signals indicative of the local magnetic environment at said measurement location(s).

19. A blast furnace according to Claim 18, wherein on or more of said magnetic sensors are moveable in said one or more hollow tubes.

20. A blast furnace according to Claim 18 or 19, further including signal processing means coupled to said magnetic sensor(s) for processing the electrical output(s) thereof.

21. A blast furnace according to any of Claims 18, 19 or 20, wherein the or each said hollow tube is provided with a cooling agent circulation system.

22. A blast furnace according to any of Claims 18 to 21, wherein a plurality of said hollow tubes are provided running through the body of said furnace at different horizontal levels below the charging level, whereby to enable detection of the passage downwardly of a boundary layer between layers of charged material having different magnetic properties.

23. A blast furnace according to any of Claims 18 to 22, wherein a plurality of the said hollow tubes are provided extending through the body of the furnace at angles to each other.

24. A blast furnace according to any of Claims 18 to 23, wherein said one or more hollow tubes include at least two tubes which extend parallel to each other in the same vertical plane of the furnace so that for each magnetic sensor located in the uppermost of the two tubes there is provided a magnetic sensor in the lower of the two tubes vertically beneath the first mentioned magnetic sensor, whereby downward movement of layers of charged material having different magnetic properties may be monitored within said

25. A blast furnace according to any of Claims 18 to 24, wherein the or each said hollow tube is double walled, comprising a first tube located within a second tube, the magnetic sensor(s) being located within the first tube(s).

26. A blast furnace according to any of claims 18 to 25, additionally including one or more charging level meters selected from sounding level meters, microwave level meters, and ultrasonic level meters, which one or more charging level meters is positioned on the body of the blast furnace for detecting the velocity of downward movement of the charge within the furnace by directly detecting the downward movement of the top surface of the packed layer of charge within a predetermined period.

27. A blast furnace substantially as hereinbefore described with reference to and as shown in the accompanying drawings.

28. A method for controlling the operation of a blast furnace, the method being substantially as hereinbefore described with reference to the accompanying drawings.

29. A blast furnace when in operation under the control of a method in accordance with Claim 11 or any claim appendant thereto or Claim 28.