JPS59226211A - Controlling method of thermal power plant - Google Patents

Abstract

PURPOSE:To enable to safely and quickly start a plant by a method wherein the rates of change of the quantities of state of the plant are set as large as the thermal stress at the turbine rotor part allows and then the steam conditions of a boiler are estimated based upon the set rates of change in order to correct said set rates of change, if the estimated steam conditions are unallowable. CONSTITUTION:At the starting of a plant, firstly, various data are inputted through a control board 10 to the plant starting initial value calculating function 210 of a computer 20 and the rates of change of turbine load, turbine speed, boiler temperature rise, boiler pressure rise and the like are set tentatively and at the same time the present temperature distribution is estimated. Secondly, the estimation time is determined at a thermal stress estimation time determining function 215 and converted to the number of changes corresponding to the change cycles of set values of the rates of change. Thirdly, the converted value is given to a steam condition estimating function 220, a packing part heat transfer rate estimating function 230, a rotor temperature distribution estimating function 240 and a rotor thermal stress estimating function 250. A rate- of-change set value deciding function 260 decides the rates of change in response to the output of the estimating function 250 and outputs said decided rates of change to a governor 270 and a boiler master 280.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a method for controlling a thermal power plant, and in particular, to a method for controlling a thermal power plant, and in particular, a method for controlling a thermal power plant to tolerate thermal stress generated in a rotor due to speed increase, temperature increase, pressure increase, and load change of a turbine. This invention relates to a control method suitable for suppressing the value below the above value and enabling safe and rapid startup.

[Background of the invention]

Recent thermal power plants are compatible with nuclear power plants,
It is starting to be operated under intermediate load operation. That is, there is a demand for starting and stopping on a daily basis, for rapid starting and stopping, or for tracking load changes. For this reason, it is important to accurately determine the thermal stress generated in the thick parts of the pressure member, and to start and stop the product in such a way that the life of the member in that part is suppressed.

One of the areas where severe stress occurs during operation of a thermal power plant is the rotor surface and pore (center hole) of the labyrinth seal after the first stage, which is exposed to high-temperature, high-speed leakage steam from the high-pressure and intermediate-pressure turbines. It is. However, since the rotor is a rotating body, it is difficult to actually measure the stress or the temperature distribution that is the basis for stress calculation.

Therefore, various operation control methods including rotor surface temperature prediction methods have been proposed. For example, the first
A method was adopted in which the post-stage casing inner wall metal temperature was regarded as the rotor surface temperature, and the stress was calculated from the actual measured values of the steam temperature and pressure after the first stage. However, with the former, there is a problem with the correlation between the casing and the rotor, and with the latter, there is a problem with measurement accuracy during no-load operation and low load when the flow rate is small. As another method, a method has conventionally been proposed in which the startup schedule is determined according to the main steam conditions immediately before ventilation to the turbine and the turbine metal temperature. However, in this method, the startup time tends to be longer than necessary because the startup schedule is created with a margin that takes into account that the main steam conditions may deviate from the scheduled values during the startup process. Therefore, with the conventional method, it is difficult to achieve a rapid start-up that effectively uses the life consumption allowed for one start-up and a rapid load change that faithfully adheres to the allowable stress. A method has also been proposed in which thermal stress is predicted during the turbine startup process and control parameters are modified based on the results.

However, in all of the above methods, thermal stress is predicted by extrapolating the values at several points in the past. Since the method was to extrapolate based on point values, it was not possible to predict the light level with sufficient accuracy. Therefore, it has been difficult to accurately control the thermal stress, which has a slow response, by proactively controlling it, and it has been difficult to accurately manage the lifetime consumption of the turbine. Therefore, it has not been possible to realize rapid startup of the turbine by effectively using the lifetime consumption allowed for one startup.

In addition, as mentioned above, in control mainly based on turbine control, although the main steam state quantity as a working fluid changes due to changes in the turbine state (speed or load), these changes are not reflected in the control. It was something that had not been done yet. That is,
Because there was no control coordination between the main steam generator (boi 2) and the turbine, it could not necessarily be said to be an optimal thermal power plant control method.

[Purpose of the invention]

An object of the present invention is to provide an optimal thermal power plant control method that can effectively utilize the lifetime consumption allowed for one startup and operation of a turbine and enables safe and rapid startup. be.

[Summary of the invention]

Wood brother E! AFi, a plant initial value setting process that temporarily sets the plant state quantity change rate of the turbine speed increase rate and load change rate, and the temperature increase rate and pressure increase rate of the turbine working fluid generator, based on the given plant startup and operation pattern. and a working fluid state quantity prediction step for predicting a change in the working fluid state quantity based on the change rate setting value and a predetermined prediction model of the working fluid generating device, and based on the working fluid state quantity prediction value. a thermal stress calculation process that predicts and calculates the thermal stress generated in the stress evaluation section of the turbine; a rate-of-change set value determining step of repeatedly executing each of the above steps to set the rate of change so that the rate of change is maximized; and controlling the corresponding turbines and working fluid generators based on the set rate of change. By adopting a control method consisting of a process,
By modifying the state quantities of the turbine and working fluid generator from time to time according to the turbine startup process and load changes, it is possible to coordinate the control of these devices, and to effectively utilize life consumption to quickly and efficiently This is to ensure safe startup and operation.

Further, the present invention makes it possible to start the turbine immediately without waiting until the temperature distribution reaches equilibrium by estimating the temperature distribution immediately before starting the turbine from the relationship between the temperature distribution when the turbine is stopped and the stop period.

[Embodiments of the invention]

Hereinafter, the present invention will be explained based on examples.

FIG. 1 is a block diagram showing the conceptual configuration of an embodiment in which the present invention is applied to a thermal power plant. In Figure 1, 1
0 is an operation panel, and 20 is a computer, which form a thermal stress prediction plant control system. Further, 30 is a coal mill system, 40 is a boiler system, and 50 is a turbine generator system.

The operator of the thermal power plant uses the operation panel 10 to perform necessary operations based on data on each part of the plant provided via the computer 20 and data provided from a higher-level control center such as a central power dispatch center (not shown). Do it above. The computer 20 provides necessary control signals to each part of the plant based on data on each part of the plant, signals corresponding to operations on the operation panel 10, etc., according to a program given in advance.

The coal mill system 30 includes a coal bunker 301 and a coal feeder 3
02, mill 310, blower 321, 322, damper 32
3,324.

Coal supplied into the mill 310 from the bunker 301 via the coal feeder 302 is pulverized in the mill, and the pulverized coal in the mill 310 is blown into the boiler by air blown by blowers 321 and 322. It is air-fed to burner 407, where it is burned.

The boiler system 40 includes a water supply tank 401. Water supply control valve 4
02, evaporator 403, -next super heater 404,
It consists of a secondary superheater 405, a gas recirculation blower 406, and a burner 407. Water supplied from the water supply pump 401 is turned into steam by an evaporator 403, and then turned into main steam, which is superheated steam, by secondary superheaters 404 and 405, and is supplied to a turbine generator system 5o, which will be described later. The amount of heat supplied by the burner 407 is used to convert water into steam in an evaporator and to superheat steam in a superheater, and a portion escapes from the chimney along with exhaust gas into the air vent. A portion of the gas exhausted from the chimney is returned to the boiler by gas recirculation pump 406. Calculator 2 is used to control the steam flow rate supplied by the boiler.
The control valve 402 is operated by the zero output. The boiler-related data for controlling the main steam temperature, such as water supply amount, secondary superheater outlet temperature, outlet steam flow rate, main steam temperature, main steam pressure, and gas recirculation flow rate, are detected by the boiler detector 411,
The signals are detected at 413, 414, 415, 416 and 417 and sent to the computer 20.

The turbine generator system 50 includes a turbine control valve 501,
High pressure turbine 502, medium and low pressure turbines 5OA, i water capacity 504, and a generator 505 directly connected to the turbine rotor
It is made up of many things. Depending on the output of the computer 20, the control valve 501
is operated, and an amount of main steam is delivered to the turbine 5 according to its opening degree.
02,503 to rotate the generator 505.

In the evening, the steam that has finished its work in the bottle is returned to water by the condenser 504. This water is again supplied to the boiler by the water supply pump 401 described above. Generator output is detected by detector 509 and sent to calculation 4320.

The computer 20 receives various requests regarding plant operation from the operation panel 10, outputs all predetermined control signals based on data obtained from the plant and a program given in advance, and controls the plant to a desired state. Immediately.

A detailed 1F5i functional configuration of the computer 20 and a sectional view of the thermal stress evaluation section are shown in FIG. As shown in FIG.
After the first stage, the labyrinth snow is opened at the mounting part 541.

The reason why thermal stress occurs when the turbine is started is that the temperature distribution inside the metal becomes uneven due to heat exchange between the metal of the turbine and the steam in contact with it. The magnitude of this heat transfer depends on the temperature difference between the metal surface and the steam and the magnitude of the heat transfer coefficient. When the turbine is started, a particular problem is the thermal stress generated in the rotor near the labyrinth/gut skin portion 541 located after the first stage of the high-pressure turbine. In other words, the temperature, pressure, and leakage flow rate of steam leaking from this area fluctuate significantly during the turbine startup process, so the rotor in this area is exposed to harsh conditions, and excessive thermal stress is generated due to rapid heating or cooling. It becomes easier to do.

Therefore, in the present embodiment, a calculation model of the boiler system 40 is used to predict the changing steam state quantity, and based on this predicted value, a predicted value of the thermal stress generated in the labyrinth part 541 is calculated. The aim is to rapidly start up the turbine and control boiler temperature and pressure increases while coordinating with boiler control.

First, the basic configuration of the control method of this embodiment will be explained.

Before starting the plant, the operator 1 inputs the turbine stop period and the plant start-up and operation pattern (hereinafter simply referred to as start-up pattern) 210 as data into the plant start-up initial value calculation function 210 via the operation panel 10. do. This function 210 temporarily sets the turbine load and speed, the rate of change of the boiler temperature and pressure increase, etc. based on the plant startup pattern, and calculates the rotor temperature return distribution measured just before the turbine is stopped, and the temperature detector 506. The current temperature distribution is estimated in consideration of the casing inner wall metal temperature 518 detected by and the rotor cooling characteristics. Next, the thermal stress prediction time determining function 215 determines how much optical thermal stress should be predicted and controlled. The predicted time at this time is based on the magnitude of the heat transfer coefficient on the rotor surface, the state of the turbine, that is, the speed 528 detected by the speed detector 507, and the load (the steam pressure after the first stage 528 detected by the pressure detector 505). (converted value from ).

Furthermore, the determined prediction time is further extended to the rate of change setting value change period (
In this embodiment, it is converted into the number of times corresponding to 3 minutes). Next, the predicted time converted into the number of changes of this rate of change set value is given to the rotor thermal stress calculation function 220, 230, 240, 250. The rotor thermal stress calculation function is composed of four functions: a steam condition prediction function 220, a packing part heat transfer coefficient prediction function 230, and a rotor temperature distribution prediction function 24010-rotor thermal stress prediction function 250. The rotor temperature distribution at the prediction time is predicted by repeating the calculation as many times as the rate of change set value change given by the function 215. Opening this temperature distribution predicted value, a function 250 calculates the entire rotor thermal stress and supplies it to the next change rate setting value determination function 260. This function 26
0 operates the thermal stress calculation functions 220 to 250 for all of the change rates prepared in advance, asks for the thermal stress tolerance value, and calculates these and the thermal stress tolerance value 207 set by the operator l.
The values that are below the thermal stress allowable value and have the maximum rate of change are set to the governor 270 as the speed increase rate set value 203 and the load change rate set value 204, and are set to the boiler master 280 as the temperature increase rate set value 205 and the pressure increase rate set value. Give as 206. In the function 260, each time the change rate set value is output to the governor 270 and the boiler master 280, the rotor temperature distribution of one period of light is initialized as the current rotor temperature distribution and used for the next prediction calculation.

Also, change the rate of change setting value to the governor 270 and boiler master 2.
Each time the target value is set to 80, the target value attainment determination function 290 determines whether the target value has been reached. If the target value is reached, the mission of this plant rapid startup system is completed; if not, the function 215 By returning and repeating the same procedure as above, the turbine is started and the boiler temperature and pressure are increased. Note that the governor 270 operates the valve position of the regulating valve so that the speed increase rate or load change rate of the turbine becomes a set change rate.

Next, details of this embodiment will be explained regarding each function.

Although it is difficult to directly measure the temperature distribution of the rotor,
Accurately determining the initial value is particularly important for this system, which aims at rapid startup. As shown in FIG. 3, the initial value calculation function 210 considers the rotor of the labyrinth packing part 541 after the high-pressure first stage to be composed of six divided coaxial cylinders, and calculates the temperature Tx-T7 of the divided surface using the following method. Calculate with. However, T1 is the rotor surface temperature and T7 is the rotor bore temperature. As shown in Fig. 4, the turbine stop period (ITB) is the rotor temperature distribution relaxation time (ts
) If there is a longer region BK, the rotor temperature is equal to the first stage rear casing inner wall metal temperature 518 actually measured by a temperature detector 506 such as a thermocouple, and the temperature distribution is considered to be uniform.

Further, when the temperature distribution is in the region A (tTs<ts) where the temperature distribution is considered to be non-uniform due to startup immediately after the turbine is stopped, the initial value of the temperature distribution is calculated assuming that the cooling characteristic of the rotor can be expressed by the following equation.

Here, T, (0) Average non-operating temperature TR when the nitrogen turbine is stopped
M: Turbine chamber temperature (constant) BM: Turbine cooling time constant C: Volume average temperature during rotor cooling Tj(o): Temperature distribution when the turbine is stopped B, : Time constant of each divided section (<BM) ( (Assuming B+rj) Next, the thermal stress prediction time determination function 215 will be described in detail with reference to FIG. 5, taking turbine startup as an example.

It is fully determined whether the current turbine state is under load operation (step 611), and the current speed (
Alternatively, the thermal stress prediction time is determined by the speed-up thermal stress prediction time determination function 612 or the load operation thermal stress prediction time determination function 613 prepared in advance as a function of the current load (L, ) (step 612 or 613). The prediction time is set as Hc L so that the value for the rotor surface and the value for the rotor bore can be determined separately, taking into account the difference in response speed of thermal stress. In this embodiment, it is determined by a linear function regarding all N, , L, and is expressed by the following equation.

If the thermal stress prediction time for the mouth surface and rotor bore are respectively △tel△1++, then during acceleration, ゝ“ (3) ∆tb (Na) = (ab bb), and during 10bb load operation, L.

△tb(L-)-”(Cb db) +db
(5)L.

becomes. However, N @ HLO is the rated speed and rated load. The thermal stress prediction time Δt1.Δtb obtained from the above equation (3) or equations (4) and (5) is the temperature distribution calculation number determination step 614 for rotor surface thermal stress prediction, and the temperature distribution calculation for rotor bore thermal stress prediction. Number of times determination step 6
In step 15, the number of calculations for each calculation is determined using the following formula % (6) (7).

Note that in equations (6) and (7), τ is the change period of the change rate setting value, and 0 is a Gaussian symbol.

Next, the steam condition prediction function 220 for determining the heat transfer coefficient essential for rotor thermal stress calculation will be explained with reference to FIG.

The function 220 is to obtain steam conditions for determining the heat transfer coefficient, that is, a second predicted steam pressure after decrepit, a first predicted steam temperature after decrepit, and a predicted value of first steam pressure after decrepit. The predicted speed value can be easily determined using the current speed value and the assumed value of the speed increase rate, and the predicted load value can be easily determined using the current load value and the assumed value of the load change rate. The predicted values of main steam temperature and main steam pressure are related to the operating state of the turbine, and it is difficult to uniquely express this correlation as a dynamic characteristic model.

Therefore, a boiler model is constructed using a method known as a Kalman filter to obtain predicted values. Figure 7 shows
FIG. 2 is a block diagram showing the basic concept of a Kalman filter.

701 is a target plant or system for which a model is to be constructed, and its dynamic characteristics are expressed by the following state transition equation.

X(++1)-(Q Lee-X(i) 10H(su・Ll(i)
-(8) Here, X(S: value of the n-dimensional state vector at time i.

Φ(Hnxn-dimensional state transition matrix.

H(i) Cn X r original drive matrix.

U (r-dimensional random variable vector representing system noise.

Assume that the noise vector U is a white random sequence, that is, the mean value and variance are given as follows.

E(Llω)−u(j)
...(9), E(W(i) - term I (u(j) - u
(j) )−δ+ jU −(10) Here, [J: r: Vector X (i), noise vector LJ (
i) Derive '. Here, it is assumed that the measuring device 703 has the following observation equation.

y(Lee C(i)・X(i)+H(Lee)
...(11) Here, y(i): m-dimensional observation vector C(i): m x n-dimensional observation matrix W(i): m-dimensional observation noise vector Observation noise vector W(i) is also a white random sequence and (9) above.
, has properties similar to those shown in equation (10). ! In addition, it is assumed that the observed noise vector W(i) is independent of the system noise vector U(i) and the initial value vector x (0). In other words, the noise vector U(i) is It is a disturbance like that.

707 is a mathematical model of the plant, and the state transition equation is expressed by the above equation (8).

Here, the state transition matrix Φ(i) is the state vector
A so-called Kalman filter is used to perform the following mathematical operations.

...(12) Here, e is the error vector, and X is the estimator of the model.

x(i)=Φ(i-1)X(1-1)+H(i-1)
u(1-1)・(13)P(1)=(M-'(i)+
C'(i)W-1C(i month-!...-(1
4) M(i)=Φ<i-1>P<i-x>Φ'(i
-1) 10H(i-1)U(i-1)H'(i-1)
...(15) 709,711 in Figure 7
are the error variance and observation matrix for calculating the above equation (12).

C of the observation matrix 711 (observation matrix C of the Rito measuring instrument 703)
We are the same.

This is a mathematical model of 713id7' runt. This model is essentially the same as Model/I/707, but instead of the maximum likelihood estimate ft That is, here j-=1.2...n, and the other conditions are the same as those for the Kalman filter described above.

Reference numeral 717 is a model correction function provided in the present invention, which constantly corrects the model in order to perform more accurate control when controlling a plant using a mathematical method known as a Kalman filter. An example of a hero in this model modified state will be explained with reference to FIG.

In Fig. 8, the horizontal axis indicates time, tl + il + 11
t and 2 respectively indicate sampling times of the observation vectors X separated by time Δt. The vertical axis is the actual value of the observation vector,
It is a predicted value. Now, as shown in equation (8) above, the state transition equation of the plant is as follows:
i) Current time 1. , the predicted value at time jl+l + tle2, i.e., xp (i, i,),
Xp (1, 2) C, respectively, are as follows.

Xp (i, 1) −Φ(')X(')+H(t
)u(i) −(19)Xp(i,'2)−Φ(
i)X (i, 1)+H(i)Ll(i) −(
20) On the other hand, the actual observed values from the plant at each time 'kX, (i +1), X, (i+2)
,! =1-ru,! =, husband's error vector e is e (i+1)=x, (i+1) -Xp(i,
1 ) −(21)e (i+2 )=X, (i+2
) -Xp(i, 2) -(22)(22)
Create -(21) and substitute (20) and (19), e(i+2) -e(i+1)=X, (i+2)
-XJ i+1 )-+Z'(i)(X(i, 1
)-X(i))...(23)...(2A) If the state transition matrix Φ(i) is optimal, e in equation (24)
(i+1)-e (i+2) should have become zero, Ruri.
Letting the state transition matrix at this time be ΦMB, this 0M8
The error vector e'(i+1) when making a prediction with
becomes as follows.

eMs (i+1)=X, (i+1)−ΦMll(
i)X(1) H(i) ・u(i)...(26) Drive matrix H (If our optimal
(i+1) should have become zero, and the total driving matrix at this time is HMII (i) and F.

That is, at each sampling of the observation vector, the actual measured value and the predicted value are compared, and the state transition matrix ΦM[+(su, drive matrix HMI+(ri) model 707, 713 By introducing the maximum likelihood estimation value x (maximum likelihood prediction value X(i, j)) with higher accuracy.

In order to solve equations (25) and (27), repeated calculations are required, which does not require a large-sized computer and also takes time. In order to simplify this, it is best to focus on the matrix that is most likely to cause errors during modeling among the matrices, and correct this. For example, it is assumed that the various heat transfer coefficients α in the state transition matrix Φ, the heat generation amount H in the drive matrix H(i), and the total amount are respectively corrected by integral calculations as follows.

α=f (e (i+2) −e (i+1)) d
t ・(28)H, =/ 6M8(i+1) d
t - (29) Although the model modification described above is based on predicted values, it can also be performed using estimated values as follows.

That is, an error vector is calculated from the estimated value and the actual measured value at each sampling time, and this is treated in the same way as the error vector in equations (28) and (29) to correct the heat transfer coefficient α calorific value H-. . A model that does not require predictive calculations is
This is better because it saves calculation time for prediction. The estimated value in this case is X, , (f −1) , X(
i) The error vector is as follows, and can be processed in the same way as the above idea.

X, (i-1)=Φ(i-2nX(i-2)+H
(i −2)u(i −2)xffl(su=Φ(i−
2)X(i-1)+H(-i-2)u(i-2)e(
i 1)=x, (t-1)-Xm(i-1)e(i)=
X, (i) -

FIG. 9 is a simplified structural diagram of the secondary superheater when the secondary superheater is considered as one lumped constant in order to simplify the model. In the figure, 421 is the secondary superheater metal tube,
422 is an outer wall. Main steam flows inside the metal tube, and all gas flows between the tube and the outer wall. Here, the following equation holds true.

From the law of conservation of energy "'(HslSn Hs2sm)°F+q2sm+
A82 s1αm82'8M (0m118H-θ5zs
n) °H' (30) where Vs2ga: Volume of secondary superheater internal fluid γ828
11: Specific weight of secondary superheater internal fluid HI! 2111
! : Enthalpy of the fluid inside the secondary superheater Fs2sn
: Flow rate of fluid inside the secondary superheater θszam 22
Next superheater outlet internal fluid temperature (steam temperature) θl11211H: Second week heater metal average temperature As25
n: Secondary superheater heat transfer area αa+l+2111f
: Heat transfer coefficient from secondary superheater metal to internal fluid = α
m52aH1r(Fs2su/Fszs4r)""...
・(31) αmgQBH,r: Heat transfer coefficient from metal to internal fluid in rated state Fs2s+i,r: Flow rate of secondary superheater internal fluid in rated state Hgtsa: Enthalpy of internal fluid at secondary superheater inlet On the other hand, secondary The following equation also holds true for the metal of the superheater tube due to the law of conservation of energy.

: As25a Hαtrn28R<θt828−θa
+a3 BH)-As2sH・α11182811(θ
l1128H-θs2sg) ・”(32) Here M+n+si: Weight of secondary superheater metal Cm+s
i: Specific heat of secondary superheater metal θz2sH: Secondary superheater external gas temperature α. 2sm:' Heat transfer coefficient from secondary superheater external gas to metal = αrgsn, r (Fgny/Fgny, r
)”...(33) 6gm28H,r: Heat transfer coefficient from gas to metal in rated state Fgnp, :Boiler gas flow rate FgBF,r: Boiler rated gas flow rate Also, secondary superheater gas temperature θg211H is θ, 2su ・...(34) Here H: : Fuel calorific value Ff : Fuel flow rate Hl : Air enthalpy F, : Air flow rate H, , : Recirculating gas enthalpy Fgrt * Recirculating gas flow rate Cpg : 1+ for gas specific heat K2 : Constant Here, when constant pressure specific heat is set as Cp = (a H/δθ)P and formula (30) is rearranged, the following formula is obtained.

Here, xl-0525m X2=θ1112811 ul:=θ[1lIIH A11= (CpFs+A+q+ga4szsu, r
(F&zsi/Fszgu,r)”8/(VB28II
・γ828H+Cp)A124s□si・αm811
sH,r (Fszsi/Fs2gu,r)”8/
(Vosi4s2sg-Cp) Bll CpFs2sm/(vszsi-vszsi
-Cp) When formula (32) is similarly generalized, it becomes the following formula.

Here, u2=θgi! 8H A21 = As2sg・α+n8JSH, r (FS2
su/Fszs+v,r)”8/(M-28HC-
211H) A22-(AS2 sn'αgm28H,r(FrBP
/FgH', r) 0°60As2I]B・α++,+
sn, r(Fszsa/F's□8H,r)"8/ (
M-zsu ・C,, zsa) B 22 = Aszsii, αgm28M,, (F
-ny/F,np,,)”/ (M-281(-C,,
, zsn) (35) and (36) are state equations representing the characteristics of the secondary superheater. Further, the secondary superheater gas temperature θ12 is determined by using the empirical formula (34) described above.

The state transition equation of the secondary superheater is expressed as X (i
+1)-Φ(j)X(i) 10H(i)u (i)
−・(37), state transition matrix Φ(
1) and the driving matrix H(t) are expressed by the following equation.

Substituting equations (35) and (36) into equation (37) and solving, ...(39) ...(40) Here, ...(41) These 't (12) ~ By introducing it into equation (15), a shicalman filter can be constructed.

In this way, the boiler system can be mathematically modeled using a Kalman filter to obtain the maximum likelihood estimated value and predicted value of the main steam temperature, but in the embodiment of the present invention, the main steam temperature can be actually measured with high accuracy. From the actual measured value and predicted value,
Modify the state transition matrix Φ (heat transfer coefficient α included in 52su, 61m21]H using the method explained using FIG. This modification enables more accurate estimation and prediction.Of course, as is clear from equations (39) and (40), the drive matrix H(i) also includes the heat transfer coefficient. , it is natural to modify these parameters in H(ri) when making the above modifications.

Furthermore, a predicted value for the main steam pressure is obtained from the boiler model in the same manner as described above. Returning to FIG. 6, when the speed prediction value and the speed increase rate assumption are given, the steam condition prediction function 622 calculates the predicted air pressure after the M2 stage (PEND), the predicted steam temperature after the first sinking (TI8T), 1. Predicted steam pressure after sunset (
PIIIT) is determined from values prepared in advance by thermal calculation, and this is corrected using predicted values of main steam temperature and pressure. In addition, the load operation steam condition prediction function 625 prepares all functions for the load so that T1117IP18T can be determined when a load prediction shift is given, and corrects it using the predicted values of main steam temperature and pressure.

Next, about the packing part heat transfer coefficient prediction function 230, the 10th
This will be explained using figures. Once the predicted values of the first steam temperature and pressure (TIIIT and PH+r), the second predicted steam pressure (P2ND), the speed predicted value (N), or the load predicted value (L>) are found, move 1 shown in FIG. The predicted value of heat transfer coefficient (K) can be obtained at T111.
? , the thermal conductivity of the steam (λIII
T>, kinematic viscosity coefficient (νIIIT), specific weight (rtg
Calculation function using steam table for T), respectively 877, 87
Find it at 8°879. Next, the packing section leakage flow rate calculation function 880 calculates the leakage flow rate (FsL) using Martin's equation expressed by the following equation: % (42): δ: Packing gap d: Rotor diameter. In addition, in the leakage flow rate calculation function 881 during load operation, the leakage flow rate is the load (L).
Assuming that it is proportional to , it is calculated using the following formula.

Feb = 0.02-p R(L/LO)
・(43) However, FR is the rated main steam flow rate, Lo
is the rated load. Therefore, the volumetric leakage flow rate (F[1L
V) is determined by the following formula. When the packing gap area is 'kA', the axial leakage flow rate ([JAX> is expressed by the following formula).

Also, the rotor surface speed (UNI) is the rotor radius r,
Then, it becomes . (45) From equation (46), composite leakage flow rate (U)
is the following formula % formula % (47). Therefore, the Raynozzle number (R,) is the rotor diameter d
Then, the Nutsselt number (N1) is expressed by the following formula.

Therefore, if the packing gap is δ, the heat transfer coefficient (K)
7 is expressed by the following formula.

Next, the rotor temperature distribution prediction function 240 will be explained.

Now consider the rotor to be an infinite cylinder, and the boundary condition λIIIT
Solve the following heat conduction equation based on , K and T18T. Here α: Temperature conductivity of rotor material r: Distance in radial direction from rotor thinness T: Rotor internal temperature t: Time The division width of the rotor is △r. Then, equation (51) can be transformed into the following differential form.

...(52) Here, in this embodiment, the change period (τ) of the change rate setting value is 3.
In the case of minutes (180 seconds), if the number of rotor divisions is 6, △(
= 0.0498 (m) and 9p In equation (52), α
τ/Δr'' = 1/2. Equation (52) can be transformed as follows.

As for the rotor surface temperature...(53) As for the rotor internal temperature, as for the rotor bore temperature. If the current time is t, the above (53) (54)
The temperature distribution after τ (seconds) is determined from equation (55). If this is repeated 2n times, the temperature distribution after nτ (seconds) can be predicted.

The above temperature distribution prediction calculation is repeated ng times for the rotor surface thermal stress and nb times for the rotor bore thermal stress.

Next, the rotor thermal stress prediction function 350 will be explained.

Based on the temperature distribution obtained by the temperature distribution prediction function 240, the volume average quality TM& and TMB at the time of predicting the thermal stress on the rotor surface and rotor bore are (T+(t+nsτ)+T+(10nsτ))...(
56) (T JCt+nbr)+T7(t-)-nhr))・
...(57) becomes. Thermal stress pre-suppression on the rotor surface and rotor bore] values σS and σb are as follows.

...(58) ...(59) where E: Young's modulus a: Coefficient of linear expansion C: Poisson's ratio and above are the thermal stress caused by the temperature distribution inside the rotor, but the centrifugal stress caused by the high speed rotation of the rotor itself Stress cannot be ignored in terms of turbine operation. If we consider the rotor as a rotating cylinder with a central hole, then the centrifugal stress (σ
. ) is expressed by the following formula.

...(60) where ρ: Rotor material density ω: Angular velocity r8 Two rotor outer radius rb: Rotor bore radius SI: Boisson's ratio Therefore, the centrifugal stress (σ.S) on the rotor surface is r = r in equation (31). In the case of g, it becomes. In addition, the centrifugal stress (σeb
) is the case where r = r h in equation (31), and it becomes. Therefore, the thermal stress described below refers to the equation expressed by the sum of equations (58) and (61) for the rotor surface, and the equation (59) and (6) for the rotor surface to taboa.
2) Shows the resultant force expressed by the sum of equations.

Next, the change rate setting value determination function 260 will be explained using FIG. 11 and FIG. 12. The optimal change rate determination function 901 shown in FIG. 11 first selects a value with a large absolute value among the rotor surface thermal stress predicted value and bore thermal stress predicted value for the change rate shown in FIG. 12. Next, if both the predicted thermal stress value and the current thermal stress value are smaller than the allowable thermal stress value set by the operator, they are stored in the value 'tII of F.

However, if the current value of thermal stress is larger than the allowable value of thermal stress, II=1 is always stored. Next, load operation judgment function 90
In step 2, it is determined whether or not it is in a load operation state, and if it is in a load operation, it is determined whether the current thermal stress calculation was for F-4, and if it is accelerating, it is determined if F=5. Determine whether or not it was. That is, it is determined whether the thermal stress prediction calculations have been completed for all the rates of change. If not, the process returns to the steam condition prediction function 220 shown in FIG. 2 for the next rate of change, that is, the rate of change corresponding to F+1, and performs thermal stress prediction calculations using the same procedure.

Furthermore, if either rate of change is inappropriate, the temperature increase rate and pressure increase rate given from the startup pattern are changed, the process returns to the steam condition prediction function 220 again, and thermal stress prediction calculation is performed using the same procedure.

In this way, the value finally stored in II is determined to be the optimal rate of change for rapid turbine startup, and is set in the governor, and at the same time, the temperature increase rate and pressure increase rate at that time are set in the boiler master. The boiler and the turbine are controlled in a coordinated manner. Here, the change rate setting functions 908, 9
After setting the change rate N or Lk number in the governor using 09,910, initialize the rotor temperature distribution with the rotor temperature distribution 3 minutes ahead corresponding to this change rate setting value and calculate the thermal stress for the next optimum change rate determination. Used as the current temperature distribution value for calculations.

As described above, according to this embodiment, based on the startup pattern, in order to effectively utilize the life consumption allowed for one startup and operation, as long as the thermal stress of the turbine rotor is allowable. The rate of change of the plant state variables is set to the maximum, and based on this, the boiler steam conditions are predicted and the rate of change is fully corrected. This allows the boiler and turbine to be controlled in a coordinated manner, resulting in high precision. This has the effect of making it possible to predict the situation and also making it possible to start up the plant quickly and safely.

〔Effect of the invention〕

As explained above, according to the present invention, it is possible to effectively utilize the lifetime consumption allowed for one startup and operation of the turbine, and moreover, it is possible to cooperatively control the working fluid generator and the turbine. Since it is possible to do this, it has the effect of being able to start up the entire town in an optimal manner, that is, safely and rapidly.

[Brief explanation of drawings]

FIG. 1 is an overall system configuration diagram of an embodiment to which the present invention is applied, FIG. 2 is a diagram showing the configuration of the main parts of the embodiment shown in FIG. 1 and the main functional block configuration of the control system, and FIG. Figure 4 is a rotor cross-sectional split diagram and rotor cooling characteristic diagram for explaining rotor temperature distribution prediction, Figure 5 is a flowchart of the thermal stress prediction time determination function, and Figure 6 is a detailed functional block diagram of the steam condition prediction function. , FIG. 7 is a block diagram showing the basic concept of the Kalman filter, FIG. 8 is a diagram for explaining model modification of the Kalman filter shown in FIG. 7, FIG. 9 is a simplified model structure diagram of the secondary superheater, and FIG. Figure 10 shows details of the rotor surface heat transfer coefficient prediction function, detailed function diagram, and diagram.
FIG. 11 is a detailed functional block diagram of the change rate setting value determination function, and FIG. 12 is a diagram showing an example of a preset change rate. DESCRIPTION OF SYMBOLS 10... Operation panel, 20... Calculator, 30... Coal mill system, 40... Boiler system, 50...
Turbine%N*'/', f”O,,,,1, Agent Patent Attorney Akio Takahashi (1,) \2, 〆/

Claims (1)

[Claims] 1. Plant initial value setting that temporarily sets the plant state quantity change rate of the turbine speed increase rate and load change rate, and the temperature increase rate and pressure increase rate of the turbine working fluid generator, based on the given plant start-up and operation pattern. a working fluid state amount prediction step that predicts a change in the working fluid state amount based on the change rate setting value and a predetermined predictive model of the working fluid generating device; and the working fluid state amount predicted value. A thermal stress calculation process that predicts and calculates the thermal stress generated in the stress evaluation section of the turbine based on the thermal stress evaluation section of the turbine, and a thermal stress calculation process that predicts the thermal stress generated in the stress evaluation section of the turbine, and that the predicted thermal stress is less than or equal to the allowable thermal stress determined based on the lifetime consumption allowed for one startup and operation. and a rate-of-change setting value determination step of repeatedly executing each of the steps to set the rate of change so that the rate of change is maximized, and controlling the corresponding turbines and working fluid generators based on the rate-of-change setting values. A thermal power plant control method characterized by comprising the steps of: