GB2250278A

GB2250278A - Sintering process employing thermal detection

Info

Publication number: GB2250278A
Application number: GB9022512A
Authority: GB
Inventors: Colin Price
Original assignee: British Steel PLC
Current assignee: British Steel PLC
Priority date: 1990-10-16
Filing date: 1990-10-16
Publication date: 1992-06-03
Also published as: GB9022512D0

Abstract

The invention is used to predict the burnthrough point for a sinter strand (9) to sinter iron ore finds before charging to a blast furnace. A series of thermocouples measure the temperature of the extracted fumes at a number of windbox (5) positions, and by manipulating these data estimates of the burnthrough point can be obtained. The invention particularly is concerned with the estimation of the transverse variation of the burnthrough point, and separate estimates of the burnthrough point are made for a number of zones transversing the strand. The invention may be used to control the charging of the sinter onto the strand so as to optimise the burnthrough point position. <IMAGE>

Description

SINTERING CONTROL The invention relates to the process of sintering, which is particularly used in the pre-treatment of iron ores before smelting in a blast furnace. As generally applied to the pre-treatment of iron ores a mixture of the iron ore together with a fuel such as coal or coke and other additives that are from time to time used is fed on to a moving strand comprising a series of grates or pallets jointed together to form a continuous band passing around rollers at either end. The charge is fed on to the horizontal upper surface of the strand at one end, ignited, often by the application of a gas flame within a hood, and is drawn slowly, by the movement of the strand over a series of laterally extending wind boxes by which air is drawn downwards through the sinter charge.The sinter charge burns downwards from the top towards the bottom as it moves along from the charging point at one end to the discharge point at the other end of the strand.

The aim for best operation of such a sinter strand is to ensure that the charge is burnt through without transverse irregularities just before the charge reaches the point at which it is discharged from the strand.

Recently, it has become a practice to divide the strand width into zones, commonly six in number, and to vary the feeding of the charge on to the strand or its composition in each zone independently, so as to be able achieve transverse regularity of the burnthrough of the charge.

Existing measurement mechanisms, such as a thermocouple measuring the average temperature of the fumes sucked through each wind box, enable detection of the average burnthrough point across the width of the strand by measuring the position of the maximum temperature, but this measurement system is obviously inadequate for the control of independent charging of the different zones. The invention has as its intention, therefore, the improved estimation of the transverse profile of the burnthrough point of a sinter strand, so that the operation of the strand is better monitored- and action can be taken, either manually or automatically, to achieve transverse regularity of the burnthrough point as well as other matters.

According to one aspect of the present invention, there is provided a method of estimating the burnthrough point in a longitudinal zone of the charge on a sinter strand, including the steps of: Arranging fixed temperature measuring devices proximate the moving charge at points in a plurality of rows spaced transversely across the width of said strand to correspond with said longitudinal zones.

For each row generating temperature profiles of the moving burning strand at least some of the points in said row.

Selecting from said profiles the temperatures of a charge element on said strand, at the time when it passed said points in said row.

Extrapolating from said temperatures the maximum temperature of said charge point and determining the charge point's position at the time when said maximum temperatures as achieved.

Outputting the said position as an estimate of the burnthrough point.

Preferably the temperature profiles are smoothed, and may be a running average. The average may be a 60 term running average.

Said extrapolation may be carried out by fitting a polynomial series to the said temperatures.

According to a further aspect of the said invention there is provided a method of validating the estimate of the position of said burnthrough point by making a second estimate of its position by means of a formula of the form BT = A - B (T(N1)/T(BTN)) where A and B are values determined experimentedly by analysis of the immediate past performance of the strand. The validation may be by checking that on the basis of the immediate past performance of the strand the estimates are statistically compatible.

One embodiment of the invention would be now described with reference to the attached figures of which: Figure la and b show in section and part plan a schematic diagram of a typical sinter strand.

Figure 2 shows the layout of temperature measurement points on a schematic plan of a sinter strand.

Figure 3 shows a typical raw temperature measurement from a thermocouple at one of the measuring points.

Figure 4 shows the smoothed temperature measurements from points T(41) and T(42).

Figure 5 shows the estimated temperature profile of a particular charge on the sinter strand as it passes down the strand.

Referring now to figure la, which is schematic only, to illustrate the operation of a typical sinter strand, grates 1 are jointed together flexibly so as to form a moveable strand 2 passing over rollers 3 and 4 by which drive may be applied. Only the horizontal upper surface of the strand is shown in Figure 1, the return path of the grates being shown in dotted outline. Below the upper horizontal surface of the strand are situated a plurality of wind boxes 6 extending laterally across the full width of the strand and through which air may be sucked by means of fans not shown communicating with the outlets from the wind boxes 5. It should be understood that the number of wind boxes shown is diagrammatic and up to 24 or more are found in actual sinter strands.A hopper 7 contains the raw sinter charge 8 comprising a mixture of iron ore, a suitable fuel such as coal or coke, and any additives that are used on a particular strand. The raw charge is discharged on to the moving strand at 9 and moves along the upper horizontal surface supported by the grates as shown at 10. A firing hood 11 commonly fuelled by gas heats the sinter charge and as it moves along in the direction of the arrow 12 air is sucked through the charge and the grates into the wind boxes as shown by arrows 13, so it begins to burn. The sinter charge burns in the form of a flame front and this is shown diagrammatically at 14, the burning starts at the top surface shortly after leaving the firing hood 11 and moves downward through the sinter charge, which is discharged at the end of the horizontal upper surface of the strand on to the receiving conveyer 15.The position of this flame front is critical to the operation of a sinter plant because the burnthrough point 16 (that is where the flame front reaches the lower surface of the charge) must be before the point at which thesinter reaches the end of the upper surface of the strand and, while this can usually be arranged, it is at the expense of, for example, an increased amount of fuel in the raw sinter charge. Also the flame front may show transverse irregularity across the width of the strand and this is illustrated in the partial plan view in figure Ib where the burnthrough point 16 is shown in plan view - as a line across the width of the strand.Clearly the average burnthrough point 17 on the strand is in advance of the last burnthrough point 18 and it will be apparent that when transverse irregularities occur the efficiency of the strand is determined by these irregularities, in that there are unused sinter strand areas.

Attempts have been made in the past to improve the transverse regularity of the burnthrough point and it is known to divide the strand longitudinally in to several, commonly six, zones which are shown from A to F in figure lb. The feeding of the raw sinter charge on to the strand at 9 in Figure la can be arranged to be independent for each of the zones A to F (this arrangement is not shown in the diagram) and thus by varying the amount, depth, or composition of raw charge fed on to the sinter strand the position of the burnthrough point in that zone can be varied. The problem is, of course, estimating where the burnthrough point in a particular zone is occurring, and indeed assembling the information as to the effects that variations in the feeding of the raw charge on to the strand have on the position of the burnthrough point in any particular zone.

Conventionally measurement of the temperatures of fumes sucked through each wind box are used to monitor the progress of the burning of the sinter.

Where the charge is burning without transverse irregularity, such a method will quite well indicate the position of the burnthrough as the wind box having the maximum fume temperature. Where there is transverse irregularity, however, because measurement of exhaust fume temperature merely gives an average across the strand, the method is not suitable for application to the control of a zoned sinter strand.

Turning now to figure 2 there is shown in schematic form a plan view last portion of the upper surface of a sinter strand 20. This part of the strand is equipped with a plurality of thermocouples arranged in a matrix to improve the estimation of the burnthrough position in each sinter strand zone.

Indicated diagrammatically are the zones A to F of the sinter strand and shown dotted lying underneath the charge and grates are the wind boxes 5. Shown are only the last eight wind boxes, and with 24 wind boxes using thermocouples in only the last eight has produced satisfactory results. Within each wind box situated as near as practicable to the underside of the grate is a line of thermocouples, six in number, corresponding to the six zones of the sinter strand.

The thermocouples will be seen fall in to rows R1, R2.... R5, R6 lying along the length of each sinter strand zone. The thermocouples may for convenience be identified by suffixes so that thermo T(ll) is the thermocouple in row 1 and wind box 1, thermocouple T(21) is the thermocouple in wind box 1 and row 2 and, for example thermocouple T54 is the thermocouple in wind box 4 and row 5. Not all thermocouples are numbered in Figure 2 to save cluttering the diagram.

Each of the thermocouples is monitored and the readings indicated by it are recorded on a computer (not shown) or other apparatus in a particular case.

Turning now to figure 3 the graph shows a typical output from a given thermocouple and this displays considerable irregularity owing to the movement of the grates, etc. It has been found advantageous to smooth the output from the thermocouples in order to use the data for further processing, and a 60 term moving average has been applied and found satisfactory. In figure 4 two such traces are shown, one from thermocouple T(41) and one from thermocouple T(42). These it will be remembered are both in thermocouples in row 4 and T(41) is the first thermocouple in that row and T(42) is the second thermocouple in that row. As the burning sinter charge moves over each thermocouple the temperature of any particular part of the moving charge in a given zone will successively be measured by the thermocouples in the particular row corresponding to that zone.In figure 4 trace T(41) and trace T(42) show the same general form, but with major trends of the trace T(42) displaced by a distance equivalent to the time taken for the charge point to move from the first of the thermocouples to the second. This distance is marked X in figure 4.

The process can be applied for all the thermocouples in a particular row so that a series of measurements of the temperature of a particular charge element as it progresses along the sinter strand zone can be obtained. The explanation will not be repeated for each individual thermocouple, since it follows exactly the same principle.

It will be apparent that as soon as any particular charge element has passed over the last wind box and has had its temperature measured by the last thermocouples in the rows its temperature as it passed over each of the thermocouples in that row can be extracted from the records and plotted, (or fed in to a computer programme to perform an equivalent calculation). These points are plotted in Figure 5 and are labelled PQRSTUVW and, as would be accepted for a normally operating sinter strand, they show a gradual increase in temperature as the charge moves along strand until it reaches the maximum temperature (where the burnthrough point is) towards the end of the sinter strand and shows a final drop in temperature over the final wind box or boxes.The position of the burnthrough point can be calculated with a little more precision, by fitting a polynomial 100 (in one of the standard manners) to the points and taking the maximum value of the polynomial as the fume temperature (T(BT)) at the burnthrough point and the position (BT) as the position of the burnthrough point. It will be remembered that this process is carried on for each of the sinter strand zones A to F and so the position of the burnthrough point in each zone is calculated, and an estimate of the transverse profile of the burnthrough points is obtained.The position as calculated can then be used either to control displays indicating the transverse profile of burnthrough point, so that manual control of the feeding of the charge on to the sinter strand can be used, or of course, the information can be fed in to an automatic control system to perform this function automatically.

An additional analysis of the data obtained from the matrix of thermocouples is possible, and can be used to provide a useful validation of the results of the method just described. The process just described produces an estimate both of the position and the fume temperature at the burnthrough point for each zone. It has been found, by experiment that a simple equation of the form BT = A - B (T(BTN)/T(N1)) where BT is the position of the burnthrough point measured from some given origin, TNl is the temperature measured by the thermocouple in the first wind box and the Nth row, and T(BTN) is the estimate of the burnthrough fume temperature for row N. A and B are numerical constants. The formula is an estimate of the burnthrough position, not as accurate as by the first method, but subject to less variation.The constants A and B depend on a particular sinter strand and its method of operation and in practice are calculated by statistical analysis of the data obtained from the matrix of thermocouples over the immediately past period of operation of the strand. An estimate also of the statistical variance of the estimate obtained by this method from the estimate of the position obtained by the first method can also be built up over that period of operation. In practice the burnthrough point for a particular sinter strand zone is measured by the first method, and is checked to be within the statistical limits of the normally operated strand based upon the immediate past performance of the strand by the second method, and if it falls within this range its value is accepted.

If it falls outside the range this is a warning the strand conditions have changed and that a check upon the estimates received is needed. Some strands may be such that a better performance is obtained if T(N2) or T(N3) is used in the equation instead of T(N1), and the best choice of reference temperature measurement is an experimental matter.

Claims

1. A method of estimating the burnthrough point in the longtitudinal zone of the charge on a sinter strand, including the steps of: Arranging fixed temperature measuring devices proximate the moving charge at point in the plurality of rows spaced transversely across the width of said strand-to correspond with said longtitudinal zones.

Selecting from said profiles the temperatures of a charge element on said strand, at the time when it passed said points in said row Extrapolating from said temperatures the maximum temperature of said charge point and determining the charge points position at the time when said maximum temperatures as achieved.

Outputting the said position as an estimate of the burnthrough point.

2. A method as claimed in claim 1 wherein the temperature profiles are smoothed.

3. A method as claimed in claim 2 wherein the temperature profiles are a running average.

4. A method as claimed in claim 3 where the average is a 60 term running average.

5. A method as claimed in any preceding claim in which the extrapolation is carried out by fitting a polynomial series to the said temperatures.

6. A method as claimed in any preceding claim further including a method of validating the estimate of the position of said burnthrough point by making a second estimate of its position by means of a formula of the form BT = A - B (T(N1)/T(BTN)) where A and B are values determined experimentedly by analysis of the immediate past performance of the strand.

7. A method as claimed in claim 6 in which the validation is made by checking that on the basis of the immediate past performance of the strand the estimates are statistically compatible.