JPS5910966B2 - How to operate a blast furnace - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a blast furnace operating method, and more specifically, it predicts with high accuracy the hot metal temperature or hot metal Si value in the blast furnace at a future time j=-, and based on the predicted value, Hot metal temperature or hot metal Si
This invention relates to a method of operating a blast furnace that enables stable operation by controlling the temperature to match a target value.

Furnace heat control is essential for stable and economical operation of blast furnaces.

That is, if the furnace heat becomes unstable, the load will drop and the ventilation state will also become unstable, which will cause changes in the bath pig iron temperature, hot metal Si value, etc., which will have an adverse effect on the subsequent steelmaking process.

Further, by stabilizing the furnace heat, it becomes possible to operate with a lower target level of furnace heat, so furnace heat control is also important from the standpoint of energy conservation.

By the way, it is said that the external factors for furnace heat fluctuations are the chemical composition and physical properties of the charging raw material.

In particular, since sintered ore accounts for the majority of the charged ore, at least 70 to 80 tons, its chemical composition and physical properties have a significant influence.

Generally, operations are carried out with consideration given to keeping the chemical composition and physical properties of the charged raw materials as constant as possible, but due to brand changes, disturbances in the sinter production process, etc.
Some degree of fluctuation is inevitable, especially when changing brands of raw materials.
Large fluctuations occur when the programming is changed.

Conventionally, the chemical components and physical properties of sintered ore have been measured and controlled at the exit of the sintering machine or sintering factory.

Therefore, the composition and properties of sintered ore that is not constantly charged into the blast furnace K:
), it is difficult to say that this is precisely understood.

In other words, a belt conveyor and a storage tank are used between the sintering machine and charging into the blast furnace. This is because there is usually a lag time of several hours to several hours between the time of measurement at the outlet and the time of charging.

Therefore, even if it is assumed that the ingredients and properties of the raw materials are the main cause of furnace heat fluctuations, it is not possible to quantify the influence on the furnace heat, or the hot metal temperature and hot metal Si value, which are indicators for follow-up fluctuations in the furnace heat control. Because of this, conventionally the reaction situation inside the furnace was estimated mainly based on the top gas analysis value,
The method of estimating the furnace heat trend by referring to past operating conditions or estimating it using a mathematical model, and controlling the fuel supply amount, etc. based on the estimation results has been adopted.

However, the response of the blast furnace to such control is slow, and it goes without saying that accurate furnace heat control cannot be achieved with feedback control based on furnace top gas analysis values, and in some cases, it is difficult to control furnace heat fluctuations due to changes in raw material composition and properties. In addition, there was a risk that the furnace heat could suddenly change due to significant changes in the ingredients and properties of the raw materials, leading to furnace cooling.

The present invention has been made in view of the above circumstances, and involves predicting in advance the influence of the chemical composition and physical properties of raw materials charged into the blast furnace from moment to moment, and performing feedforward control based on the results of this prediction. It is an object of the present invention to provide a method of operating a blast furnace that enables stable and economical operation by stabilizing furnace heat with high precision.

The blast furnace operating method according to the present invention includes actually measuring the chemical composition and physical properties of the raw material to be charged into the blast furnace over time from the raw material manufacturing factory to the blast furnace,
In addition, by tracking the movement of raw materials during the transportation process up to the blast furnace, we can determine the chemical composition and physical properties of the raw materials that are being charged into the blast furnace moment by moment, and the chemical composition and physical properties of the raw materials that have been determined in advance. Based on the response characteristics of hot metal temperature or hot metal Si value to changes in hot metal Based on the response characteristics of the reaction rate in the furnace to changes in the manipulated variable, the reaction rate in the furnace and the temperature inside the furnace at future times can be calculated from the top gas analysis values obtained every moment and the response characteristics of the reaction rate in the furnace. From this predicted value and the actual measured value of hot metal temperature or hot metal Si value, predict the hot metal temperature or hot metal Si value at a future time, and use this predicted value to predict the amount of change in the hot metal temperature or hot metal Si value. The method is characterized in that a corrected predicted value of the hot metal temperature or the hot metal Si value obtained by adding the values is obtained, and the hot metal temperature or the hot metal Si value is controlled according to the following equation.

However, U*: Manipulated amount after change UO ++. Current time manipulated variable GT3: Iron X*: Target value of hot metal temperature or hot metal Si value X,: Hot metal temperature or molten metal that will rise at future time j Corrected predicted value of pig Si value The method of the present invention will be explained in detail below with reference to the drawings.

FIG. 1 schematically shows the route of the blast furnace raw materials such as sintered ore and coke to the blast furnace 14.

Raw materials such as sintered ore are available at 10 sintering plants and 10 coke manufacturing plants.
' etc. are carried out by the belt conveyor 1, stored in the respective storage tanks 12, and then cut out depending on the unloading situation, transported by the charging conveyor 3, and loaded into the blast furnace 14 from the top of the furnace. It is designed to be entered.

Now, how to determine the chemical composition and physical properties of the raw material charged into the blast furnace 14 without delay, that is, the composition and physical properties of the raw material actually charged into the blast furnace at a certain point in time.
We will explain how to obtain the properties.

The basic principle of this method is that the ingredients and properties are first sampled and measured at appropriate intervals at a raw material manufacturing factory such as a sintering factory, as in conventional methods, and the transportation process of the raw material of the sampled portion is controlled by a computer and charged into the blast furnace. We track the ingredients and properties of the sampling target area and the raw materials before and after it.
The sampling target portion is sequentially replaced with the properties of the components obtained by measuring them or their modified data.

Although the measurement results of the components and properties are known later than the sampling time, there is no problem in that it takes several hours or more before the sampled part is charged into the blast furnace, as mentioned above.

Hereinafter, assuming that there is one ore storage tank for sintered ore, an explanation will be given by taking as an example RDI (reduction powder property), which is one of the physical property measurement items.

First, in the conveyance process by the belt conveyor 11 from the sintering factory to the ore storage tank 12 and the conveyance process by the charging conveyor 13 from the ore storage tank 12 to the blast furnace 14, the time required for each process is approximately constant. Tracking in these processes can be easily performed by processing data using the required time as a time delay.

Of course, the conveyance speed of the belt conveyor 11 and the conveyance speed of the charging conveyor 13 may be input into the computer over time to improve the tracking accuracy.

Next, a tracking method within the ore storage tank 12 will be described.

As shown in FIG. 2, the ore storage tank 12 has a constant volume Δv (rrL3
) is recognized divided in the height direction into a number of thin horizontal layers,
The following calculations are performed at each timing at the end of charging into the ore storage tank 12 and at the end of cutting from the ore storage tank 12, and each horizontal layer R
Let's find DI.

The meanings of the symbols used below to explain this operation are as follows.

The horizontal layer numbers are lI 2 - 3... from the bottom.
As you move upwards, it becomes a dog.

Nmax: Number of the uppermost horizontal layer in which sintered ore exists Nmax: NmaxO value ρ immediately before charging or cutting
(#g/door): Bulk density of sintered ore Win (#g): Charging amount Wout (#9): Cutting amount β: Sintered ore filling rate in the uppermost horizontal layer (for example, the second
β: The value of β immediately before charging or cutting (for example, the ratio of the small dot in Figure 2 to ΔV) RDIO value trasoked per layer RDIjφ) charging or cutting in horizontal layer of second number j △
Immediately before output 0 f') RDI (7) Value RDIk (→: R of sintered ore charged in the horizontal layer with number k
DIO value tins: Time of charging start tinE: End of charging tin (Nmax+i): Time W when sintered ore is charged into the horizontal layer with number Nmax+i, (kV minutes): From sintering factory to ore storage tank Belt conveyor conveyance speed T8 (minutes): Required time for conveyance from the sintering factory to the ore storage tank TB (minutes): Required time for conveyance from the ore storage tank to the blast furnace Nin
: Number of horizontal layers increased by charging Nout : Number of horizontal layers decreased by cutting α : Volume ratio of the remaining sintered ore in the uppermost horizontal layer involved in cutting to Δ■ (for example, a) At the end of charging Figure 2 schematically shows the state inside the ore storage tank at the end of charging, and the two-dot chain line in the figure indicates the level l just before charging It shows.

Referring to this figure, it is clear that the relational expression holds true.

On the other hand, the RDI value for each horizontal layer of charged sintered ore (indicated by the diagonal line downward to the left in the figure) is different from the RDI average value in the above-mentioned time interval, for example, (tin(Nmax+i
-1) *tin(Nmax+i),
To explain the RDI average value within a one-hour period, it is a value obtained as follows.

That is, in FIG. 3, time is plotted on the horizontal axis, and RDI value is plotted on the vertical axis, and the RDI measurement results and sampling time points are shown in correspondence with each other by circles.

Then, tin{Nmax+i-1) and tin(Nmax
x+i)) (i.e., the time when the sintered ore charged into the ore storage tank at each time passed the sampling position of the sintering factory) tin(Nmax+
i −1 )−′rg and tin(Nmax+i)−T
8 can be represented, for example, by two vertical lines in the figure, the RDI average value is determined as the average value (indicated by a broken line) of the linearly interpolated values of the ○-marked coordinates within this vertical line section. .

In this way, RDIk (here k=Nmax~Nm
ax), the RDI in each horizontal layer in the ore storage tank can be calculated using the following formula.

(Opening) At the end of cutting, Figure 4a shows the state before the cutting value. b. are schematic diagrams of the conditions inside the ore storage tank after quarrying. It shows.

The diagonally shaded part on the lower left side of the figure is the cut out part.

Referring to this figure, each horizontal layer KJ in the ore storage tank is calculated based on the value of J Saishin.
Jd RDI can be calculated using the following formula.

In this way, the RDI value of the sintered ore in the ore storage tank will be tracked.The RDI of the horizontal layer that affects the force ζ cutting is RDI. (j-"l-2=Nout), the average RDI of one charge of sintered ore charged into the blast furnace is calculated as (cutting time + TB
) Tracking is performed assuming that the material is charged into the blast furnace at the time.

I have described the RDI tracking of sintered ore above, but
The same process can be applied to other physical properties or chemical components of the sintered ore, and even to the components and properties of other raw materials such as coke.

If multiple ore storage tanks are prepared for each raw material, it goes without saying that tracking is performed for each ore storage tank, but the property values tracked for each tank are weighted by the amount cut from each tank. An average value may be used.

Although the components and properties of raw materials cannot generally be measured continuously or in real time, it is possible to measure the RDI mentioned in the above explanation continuously and in real time.

That is, it is known that there is a strong correlation between the magnetic permeability of sintered ore and RDI.

Therefore, as shown in FIG.
By placing it at the exit of the charging conveyor 13, etc., and continuously measuring the magnetic permeability of the passing sintered ore,
Indirect RDI can be measured continuously and in real time.

Therefore, if the time required for transportation from the location of the magnetic measuring device 15 to the blast furnace (approximately constant) is given, tracking can be performed extremely easily as far as RDI is concerned.

Now, the step response of the hot metal temperature to changes in the composition and properties of the raw material is determined in advance by a blast furnace step response experiment or by analysis of blast furnace data.

That is, as shown in Figure 5a, the components or properties (those selected as measurement items because they affect the furnace heat).

For example, RDI of sintered ore, iron content, rotational strength, average particle size, cold strength of coke, hot strength, average particle size, etc.) Un
changes stepwise by ΔUn, the fifth
As shown in Figure b, the fluctuation amount Tpig of the hot metal temperature Tpig which changes gradually with a time delay is determined.

Then, the response coefficient up to time T when the response is considered to reach a steady state is determined using the following equation (16).

Figure 5C shows the results.

In addition! K! indicates the discrete time that is the horizontal axis coordinate value of . , Tme and U! mentioned below. nplg
n, etc. means the value at that time.

Now, when K7' is determined in this way, by using the changes in U that are known through tracking, the temperature T of hot metal caused by changes in the ingredients and properties of the raw material
・The change in can be predicted.

That is, the amount of change ΔTpig in Tpig in j future time months, which is caused by the components and properties, is expressed by ΔTpig.

Note that Σ means taking the sum of the calculation results in parentheses for all items of ingredients and properties.

Note that here, the step response of the hot metal temperature was determined in advance, and the amount of change in the hot metal temperature at time j was predicted from this and the tracking results, but the step response of the hot metal Si value was determined, Similarly, the amount of change at time j may be predicted.

In this way, the furnace heat fluctuations due to changes in the ingredients and properties of the raw materials are predicted, or in other words, the amount of change in the hot metal temperature and hot metal Si value at a future time is predicted.

On the other hand, the prediction of the furnace heat transition based on the current furnace heat estimated from the blowing conditions and the top gas components and the past operation amount has already been completed and proposed by the inventor of the present invention in Japanese Patent Application No. 10600/1983.
This can be carried out according to invention No. 3.

First, the mathematical model used for current estimation and future prediction of furnace heat will be explained.

As shown in Figure 6, the mathematical model consists of a preheating zone (first layer) at the top of the furnace, a Fe 2 Q 3 reduction zone (second layer), and a Fe 2 Q 3 reduction zone (second layer).
3Q4 reduction zone (third layer), FeQ reduction and direct reduction reaction zone (fourth layer), and carbon combustion zone (fifth layer).
This is a mathematical model that is divided into two layers and formulas for material, pressure, and heat balance are established for each layer, and the distribution of materials and movement of materials inside the furnace are expressed as shown in Figure 7.

The procedure for calculating the current furnace internal temperature is based on the current operating variables, that is, the tuyere operating variables (air flow rate, enriched oxygen, moisture, air temperature, liquid fuel), ore/coke, and top charge composition. , the reaction rate R1 to R1o (see Figure 6) is calculated from the top gas composition, the mass transfer flow rate at each stage is calculated from R1 to Rlo, and the heat balance equation (first-order differential equation) is solved to determine the solid temperature at each stage. Calculate TSi and gas temperature TGi.

The reaction rates R1 to RIO are quantities defined by the reaction equation shown in FIG. 6, and the current times R1 to Rto are determined by the following equation.

(R6*R7*RB can be found immediately from the air blowing conditions)
(R 4 1 R 5 * R 9 is determined as follows using the furnace top gas analysis value) (Furthermore, the following formula holds true during operation while maintaining the charging speed at a constant stock line) However, the reaction rate R1 to R 'to is determined, the amount of mass transfer at each stage (
S) i, gas movement amount (C) i is (27). It is found using (28).

however Furthermore, S.

can be calculated from R5 and the charge composition under constant stock line operation.

Further, G5 can be determined from the air blowing conditions.

Once the moving flow rate of each stage is determined, the heat balance equation for each stage (29) (3
0) at the current time, TS i * TG
t is found.

Here, the procedure for calculating the furnace internal temperature at the future time is to apply exactly the same calculation method as the current time estimation explained earlier, except for predicting the reaction rate at the future time, and solving the mass transfer flow rate heat balance equation. do.

The reaction rate prediction method that is characteristic of this prediction method is as described below.

That is, in the same way as the method of determining the responsiveness to changes in the ingredients and properties of raw materials as explained based on Fig. 5, first, the manipulated variable Un (n: heavy oil, air flow rate) is determined by blast furnace data analysis or step response experiment, etc. , enriched oxygen, charging coke ratio, blast temperature, moisture, furnace top pressure)
4. Examine the response of R5 and determine the response coefficients K7 and K4 of the reaction rate equation expressed by equations (3l) and (32).

n n From these, RA at time j based on the operation change amount ΔUn. R
Y is the following formula (33). It becomes like (34).

Next, as explained earlier, since the current reaction rate can be calculated from the f1 top gas composition, the
Using j * K'j, the reaction rate equation is expressed as (35). (
36).

Regarding other reaction rates, (37) to (43)
Perform predictive calculations using formulas.

After predicting the future reaction rate as described above, the material flow rate calculation and heat balance calculation are performed, and the internal temperature at the future time is T81%.
Predictively calculate the gas temperature TGi.

Among the calculated values mentioned above, the calculated furnace lower solid temperature TS5 corresponds extremely well to the actual hot metal temperature and the actual hot metal Si value, proving that it is possible to predict the hot metal temperature or hot metal Si value with high accuracy. ing.

The calculated solid temperature TS5 and the actual hot metal temperature Tpig or the actual hot metal Si value correspond well, but in the long term, due to drift in the measured value and changes in blast furnace heat loss, TS5 and T, j,
X may cause a difference in V pel from Si.

Therefore, in order to control the bath iron temperature or hot metal Si value, it is necessary to appropriately correct the level difference.

For example, to explain the case of controlling using the hot metal temperature as an index, the predicted hot metal temperature is calculated using the difference δTpig between the measured hot metal temperature Tpig and the calculated current furnace lower temperature TS5 at the measurement time, as shown in equation (44) (45). Fix it.

(See FIG. 8) However, for δTpig, an average value of several taps can be used to eliminate the influence of measurement errors.

This is how I got it! represents the predicted value of the hot metal temperature at time j when the manipulated variable of the base time is maintained into the future.

In this way, the predicted hot metal temperature T ノ at future time j is obtained using equation (44), but in the present invention, the predicted value of the change in hot metal temperature at future time j based on the ingredients and properties of the raw material, that is, (l
7) It is corrected by taking into account ΔTJ expressed by the formula, and the temperature of the hot metal at that future time j is estimated.

The manipulated variable for stabilizing the furnace heat is the corrected predicted value of the hot metal temperature as described above, that is, TJ, +ΔT. Based on the deviation between j and the target hot metal temperature Tt, it is determined by the ptg formula.

However, as the manipulated variables to be changed, the manipulated variables determined by the operating policy can be arbitrarily selected from among the air blowing temperature, humidity, heavy oil injection amount, ore/coke, etc.

FIGS. 8 and 9 are conceptual diagrams of the operating method according to the present invention, taking the amount of heavy oil injection as an example.

That is, when the amount of heavy oil injection among the various manipulated variables is changed as shown in Figure 8A, it is necessary to further change the amount of heavy oil injection in order to control the hot metal temperature at the future time to the target value. Determine the amount.

This is done by calculating the fluctuation amounts ΔR and ΔR of R and R using equations (35) and (37) from the previously determined KIK'' [Fig. 91g [Fig. 8C], and furthermore, this hot metal temperature Tp'i g
Assuming that the temperature corrected by adding ΔTp1ig to is a more accurate future hot metal temperature, the amount of change in the heavy oil injection amount at the current time is determined according to equation (46) so that this future hot metal temperature becomes the target temperature. It is.

In Figure 9, (Formula 460 is shown above,
It is expressed as r 1 0, and is substantially the same as (46)2.

In this way, the present invention focuses on the chemical composition and physical properties of charging materials, which have been considered to be one of the main factors of furnace heat fluctuations, but whose effects could not be systematically reflected in the analysis of furnace heat fluctuations. As for the properties, by tracking the raw material transportation process, we obtain the momentary values and predict the furnace heat fluctuations based on the changes in advance, and further predict the furnace heat fluctuations due to other factors by separate means. Based on the results of both predictions, more accurate predictions of furnace heat fluctuations are made, which enables highly accurate furnace heat control, and is extremely effective in equalizing pig iron grade and saving energy by reducing the target heat level. It is.

In the embodiments described above, the hot metal temperature was used as an index for furnace heat control or monitoring, but the present invention can be carried out in exactly the same way even if the hot metal Si value is used instead.

[Brief explanation of the drawing]

Figure 1 is a schematic diagram showing the raw material conveyance system, Figure 2 is a conceptual diagram showing the state of the ore storage tank after raw material is charged, Figure 3 is an explanatory diagram conceptually showing how to calculate the RDI average feed rate, Figures 4a and 4b are conceptual diagrams showing the state of the ore storage tank before and after raw material cutting, respectively.
Figures A, B, and C are graphs showing the time trends of Fufu Un, Tp, g, and K《, Figure 6 is an explanatory diagram showing a reaction model inside the blast furnace, and Figure 7 is the inside of the blast furnace shown in Figure 6. Figure 8 is an explanatory diagram of a method for predicting the hot metal temperature at a future time from the predicted TS at a future time and the actually measured hot metal temperature. Figure 9 is a schematic diagram showing the distribution and movement of materials at a future time. FIG. 7 is an explanatory diagram showing a method of determining the amount of change in the operation amount necessary for control from the corrected predicted value. 10... Sintering work 411... Belt conveyor, 12... Ore storage tank, 13... Charging conveyor, 14... Blast furnace , 15...Magnetic measuring device.

Claims (1)

[Claims] 1. The chemical composition and physical properties of the raw material to be charged into the blast furnace are actually measured over time from the raw material manufacturing factory to the blast furnace, and the raw material is By tracking the movement, the chemical composition and physical properties of the raw material being charged into the blast furnace are determined moment by moment, and the response of hot metal temperature or hot metal Si value to changes in the chemical composition and physical properties of the raw material, which have been determined in advance, is determined. Based on the characteristics, while predicting the amount of change in the hot metal temperature or hot metal Si value due to changes in the chemical composition and physical properties determined as described above, Based on the response characteristics of the furnace top gas analysis values obtained every moment and the response characteristics of the furnace reaction speed, the furnace reaction rate and furnace internal temperature at a future time are predicted, and this predicted value and the hot metal temperature or The hot metal temperature or hot metal Si value obtained by predicting the hot metal temperature or hot metal Si value at a future time from the actual measured value of the hot metal Si value, and adding the predicted value of the amount of change in the hot metal temperature or hot metal Si value to this predicted value. A method for operating a blast furnace, comprising: obtaining a corrected predicted value, and controlling the hot metal temperature or the hot metal Si value according to the following equation. However, U*: Post-change manipulated variable UO: Current time manipulated variable GJ: Constant X*: Target value of hot metal temperature or hot metal Si value Xj: Corrected predicted value of hot metal temperature or hot metal Si value at time j