JPH0480282B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a method for monitoring thermal stress of a boiler, and particularly to improving the accuracy and reliability of thermal stress monitoring and life management.

[Background of the invention]

When the boiler starts, stops, or changes in load, the fluid temperature fluctuates greatly and a large difference occurs between the fluid temperature and the pressure member temperature. This generates thermal stress in the pressure member, especially in the nozzle corner of a thick pressure-resistant part such as the header header at the outlet of the secondary superheater.
This results in large thermal stress, which greatly reduces the fatigue life of the pressure member.

Therefore, from the perspective of preventing damage to pressure members and ensuring safety in boiler operation, it is desired to suppress the life consumption of pressure members due to boiler startup, shutdown, load changes, etc. to below a certain allowable value. ing. To this end, it is extremely important to accurately understand the life consumption of pressure members. It is necessary to accurately understand thermal stress.

Now, this thermal stress is determined based on the temperature difference between the average temperature of the pressure member (hereinafter referred to as metal) and its inner surface temperature, but these temperatures are usually calculated based on the temperature distribution in the thickness direction of the metal (hereinafter referred to as metal). (abbreviated as metal temperature distribution). Therefore, in order to accurately understand thermal stress, it is important to accurately determine the metal temperature distribution.

To obtain the metal temperature distribution, temperature boundary conditions must be set for the inner and outer surfaces of the metal. Conventionally, the method of setting this condition was
As shown in No. 58-52563, (1) metal temperature detectors are attached to the inner and outer surfaces of the stress monitoring evaluation points set on the pressure member, and each actual value detected by this is measured at the temperature boundary. (hereinafter referred to as the second method); or (2) detect the temperature of the internal fluid, estimate the heat transfer coefficient on the inner surface of the metal based on this measured value, and A method of setting temperature boundary conditions (hereinafter referred to as the first method) has been proposed. The former method has high accuracy in calculating the metal temperature distribution because it directly measures the temperature of the metal, but it requires a temperature detector to be embedded in a high-temperature part.
There is a high probability that the temperature detector will break, and the reliability of the stress monitoring system is low. On the other hand, in the latter method, the process state quantities such as internal fluid flow rate, pressure, and temperature detected by the detector installed in a normal boiler automatic control system can be used as they are, so the temperature sensor There is no need to provide a separate one. Therefore, since there is no decrease in reliability due to an increase in the number of detectors, the first method is more advantageous, especially in a region where a sufficient flow rate of the internal fluid is ensured and the heat transfer coefficient is stable. However, according to this method, the heat transfer coefficient is generally determined using experiments or theoretical formulas based on a steady state, so when the state quantity of the internal fluid is fluctuating, the heat transfer coefficient cannot be calculated. There was a fear that the calculation accuracy would deteriorate and the results would be far different from the actual values. Furthermore, since there was no effective method to check the calculation accuracy, there was a drawback that the accuracy of the determined metal temperature distribution and, by extension, the value of thermal stress were not guaranteed.

[Purpose of the invention]

An object of the present invention is to provide a boiler thermal stress monitoring method that can accurately determine thermal stress and has excellent reliability.

[Summary of the invention]

The present invention detects the state quantity of an internal fluid at a stress monitoring evaluation point set in a boiler member, and estimates the heat transfer coefficient of the interface between the internal fluid and the member at the evaluation point based on this detected data. a first method of determining the temperature distribution in the thickness direction of the member by determining the amount of heat input to the member based on the estimated heat transfer coefficient and the detected data; and a second method for determining the amount of heat input to the member and the temperature distribution in the thickness direction of the member based on the inner surface temperature and outer surface temperature; the temperature distribution obtained by the first method and the second method. and when the difference is greater than or equal to the allowable value set based on the standard of whether or not the difference in the corresponding stress values is allowable, the heat input of each of the above methods corresponding to each temperature distribution is compared, and the When the heat input amount of the first method is larger than the heat input amount of the second method, the first method
The estimated value of the heat transfer coefficient by the method of 2 is reduced, and when it is small, the estimated value of the heat transfer coefficient is increased. Let us calculate the stress,
By modifying the first method, which has a low probability of detector failure, using the second method, which is more accurate, as a standard, the accuracy of thermal stress is improved and reliability is improved.

[Embodiments of the invention]

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

FIG. 1 shows a schematic configuration of a boiler thermal stress monitoring device 100 to which the present invention is applied and a boiler 200 to be monitored. As shown in the partially enlarged view in the same figure, the evaluation points of the thermal stress monitoring in this example were based on the inner nozzle corner of the secondary superheater outlet header header 202, which is under severe conditions for both thermal stress and life consumption. Part NC is set. In other words, since high-temperature steam of approximately 550° C. flows into this portion from the secondary superheater 201, a large temperature difference occurs between the inner and outer surfaces of the metal, and this causes a large thermal stress to act.

In FIG. 1, a boiler thermal stress monitoring device 100
are a digital computer 101 and an alarm display 10.
2, CRT display device 103, typewriter 104,
and a hard copy device 105.
The digital computer 101 calculates secondary superheating based on the main steam pressure P, the main steam flow rate G, the main steam temperature T _f , and the inner metal temperature T _M1 and outer metal temperature T _M0 of the secondary superheater outlet header header 202. The thermal stress and lifetime consumption of the nozzle corner NC of the header header 202 are calculated at a predetermined cycle, and the thermal stress is displayed on the alarm display 102 at the time of alarm.
While displaying related information on the CRT display device 103,
Print it on the typewriter 104. Also,
The CRT display device 103 displays trend graphs and numerical displays of calculated thermal stress values, life consumption data per start/stop, and cumulative life consumption data. The data is displayed and printed according to the operator's request. Information such as trend graph display displayed on the CRT display device 103 is recorded by the hard copy device 105 upon request.

FIG. 2 shows the processing functions of the boiler thermal stress monitoring device 100 in blocks.

First, the second metal temperature calculation section 10
It is started by the start command from the start circuit 11, and the metal inner surface temperature detection value is
The metal temperature distribution is calculated using T _M1 and the metal outer surface temperature detection value T _M0 as temperature boundary values.

On the other hand, the first metal temperature distribution calculation unit 12 corresponds to the first method, and the temperature T _f of the internal fluid,
Using the heat transfer coefficient estimated by the heat transfer coefficient calculating section 13 based on the pressure P and the flow rate G, the amount of heat input from the internal fluid to the metal is determined, and the metal temperature distribution is calculated. The heat transfer coefficient correction unit 14 compares the calculation results of the first and second metal temperature distribution calculation units, and when the difference is greater than the allowable value, the heat transfer coefficient calculation unit 13 performs calculation to reduce the difference. The estimated heat transfer coefficient is corrected, and the details of the process will be described later. The diagnostic circuit 15 is a part that detects an abnormality in the metal temperature detector, and when the detector is abnormal, the correction lock circuit 1 is activated.
A lock signal is sent to prevent the output of 6, thereby stopping the heat transfer coefficient correction operation based on the abnormal metal temperature distribution calculated based on the abnormal metal temperature detection value.

The thermal stress calculation section 17 calculates thermal stress based on the metal temperature distribution calculation value outputted from the first metal temperature distribution calculation section 12 by the method shown in Japanese Patent Application No. 58-52563. There is. Based on this thermal stress, the life consumption calculation unit 18 calculates the
-223939, the lifetime consumption is calculated.

The alarm processing section 19 compares the alarm setting value outputted from the alarm setting section 20 with the calculated value of thermal stress, and displays an alarm on the alarm display 102. The CRT display processing unit 21 displays thermal stress calculation results, related data, and life consumption calculation results.
Processing for displaying on the CRT display device 103 is performed. The typewriter 104 is configured to print alarm-related data when an alarm occurs, thermal stress calculation results, data such as life consumption.

Next, a method of calculating metal temperature distribution in the second metal temperature calculating section 10 will be explained.

Assuming that the secondary superheater outlet header header 202 is an infinite cylinder as shown in FIG. 3, and that no temperature distribution occurs in the axial direction, the temperature distribution within the metal is expressed by the following equation.

1/α′ ∂T/∂t=∂2T/∂r ² +1/r ∂T/∂r…
…(1) T (a, t) = T _Mi (t) … (2) T (b, t) = T _Mp (t) … (3) However, α′: Thermal diffusivity T: From the center Radius r, metal temperature at time t T _Mi : Metal temperature measurement value at time t on the metal inner surface (r=a) T _Mp : Metal temperature measurement value at time t on the metal outer surface (r=b) Equations (1) ~ ( 3) is calculated using a digital calculator,
As shown in Figure 4, the metal is divided into N parts in the radial direction, and the metal temperature T ₁ (j) at each division point is determined.
It is converted into a differential format with the division unit length as Δr and the temperature distribution calculation period as Δt.

The state of this conversion is shown in FIG. In FIG. 5, 70 is a table storing data from temperature T _M1 (j) near the inner surface of the metal to temperature T _M0 (j) of the outer surface of the metal. Note that (j) for other data hereinafter indicates data at time j. 71 is a conversion unit that inputs this data and converts it into a differential format. 72 is a table that stores the conversion results of the conversion unit 71.

70, T _M1 (j), T ₁ (j), T ₂ (j)...T _k-2 (j),
Data of T _k-1 (j), T _k (j), T _k+1 (j), ... T _N-1 (j), T _M0 (j) are stored. Here, T _M1 (j) is the temperature near the inner surface of the metal at time j, T _Mp (j) is the temperature of the outer surface of the metal at time j, and T ₁ (j) is the dividing point i of the metal (i=1,
2,...k-2,k-1,k,k+1,...N-
1) shows the temperature at time j. This data 70 is converted into a differential format by a conversion unit 71.
That is, T ₀ (j+1)=T _M1 (j) BT ₀ (j+1)−AT ₁ (j+1)+T ₂ (j+1) =−BT ₀ (j)+CT ₁ (j)−T ₂ (j) 〓 BT _{k -1} (j+1)−AT _k (j+1)+T _k+1 (j+1) =−BT _k-1 (j)+CT _k (j)−T _k+1 (j) 〓 −B/AT _N-2 (j+1 )−T _N-1 (j+1)−1/AT _N (
j +1) =-B/AT _N-2 (j)-C/AT _N-1 (j)+1/AT _N (j). However, A={1/(Δr) ² +1/α _j Δt}/1/2Δr(1/Δr)
r+1/2j) B=1/2Δr(1/Δr+1/2r ₁ )/1/2Δr(1/Δr
r+1/2r ₁ ) C={1/(Δr) ² +1/α _j Δt}/1/2Δr(1/Δ
r+1/2r ₁ ) T _M1 (j): Temperature near the inner surface of the metal at time j T _M0 (j): Temperature of the outer surface of the metal at time j T ₁ (j): Temperature at time j at dividing point i of the metal Δt: Sampling period α _j ′: Thermal diffusivity Δr: Unit length of metal division in the radial direction r ₁ : Distance from the center of the cylinder The results of this conversion are stored as a table 72.

Next, the method of metal temperature distribution calculation in the first metal temperature distribution calculation section 12 is explained in the heat transfer coefficient calculation section 1.
This will be explained in conjunction with 3. In this part, the temperature distribution on the inner surface of the metal is expressed by the following equation.

1/α′ ∂T/∂t=∂2T/∂r ² +1/r ∂T/∂r…
…(4) −λ・∂T/∂r｜ _rn 〓=h(T _f −T ₀ …(5) T(b, t)=T _M0 (t) …(6) However, λ: Metal Heat transfer coefficient h: Heat transfer coefficient T _f Internal fluid temperature = T _f (t) T ₀ : Calculated value of internal metal temperature Formulas (4) to (6) are calculated in the same way as in the second metal temperature distribution calculation section 10. When converted into the differential format, it becomes as shown in Fig. 6. In Fig. 6, 80 is a table that stores the data before conversion, 81 is a conversion unit that converts the data into the differential format, and 82 is the data after conversion. The data stored in 80 is converted by the conversion unit 81 as follows: -B・D/A+B・DT _f (j+1)+T ₀ (j+1)-1
+B/A+B・DT ₁ (j+1) =B・D/A+B・DT _f (j)−C+B・D/A+B・DT
₀ (j)+1+/A+B・DT ₁ (j) B・T ₀ (j+1)−A・T ₁ (j+1)+T ₂ (j+1) =−BT ₀ (j)+C・T ₁ (j)−T ₂ (j) B・T _k-1 (j+1)−AT _k (j+1)+T _k+1 (j+
1) =−BT _k-1 (j)+C・T _k (j)−T _k+1 (j) B・T _k (j+1)−A・T _k+1 (j+1)+T _k+2 (j+1) =-B・T _k (j)+C・T _k+1 (j)−T _k+2 (j) −B/AT _N-2 (j+1)+T _N-1 (j+1)+1/AT _M0
(j +1) =B/AT _N-2 (j)-C/AT _N-1 (j)+1/AT _M0 (j) Here, A={1/(Δr) ² +1/α _j ′ Δt}/1/2Δr(1/
Δr+1/2r ₁ B=1/2Δr(1/Δr−1/2r)/1/2Δr(1/Δr
+1/2r ₁ ) C={1/(Δr) ² +1/α _j ′Δt}/1/2Δ−(1
/Δr+1/2r ₁ ) D=2h _j δr／λ _j T _f (j): Temperature of internal fluid at time j T ₁ (j): Temperature of dividing point i of metal at time j T _M (j): Metal Temperature of the outer surface at time j α _j ′: Thermal diffusivity λ _j : Heat transfer coefficient h _j (j): Heat transfer coefficient Δr: Unit length of metal cylinder division r ₁ : Distance from the center of the cylinder Next, heat transfer The rate h _j is calculated based on the following formula.

W _j = G _j・v/S …(7) S=π/4a ² …(8) R ₀ =W _j・2a/ν …(9) h _j =0.023・R ₀ ^0.3・P _r ^0.4・K/2a ...(10) where W _j : Steam flow rate G _j : Internal fluid flow rate v : Specific volume (function of temperature T _f and pressure P) S : Channel cross-sectional area R ₀ : Reynolds number ν : Kinematic viscosity coefficient (function of temperature T _f and pressure P) K: Heat transfer coefficient (function of temperature T _f and pressure P) P _r : Prandtl number (function of temperature T _f and pressure P) a: Inner radius Then, based on the metal temperature distribution calculation values obtained by the first and second metal temperature distribution calculation units 10 and 12, the heat transfer coefficient correction unit 14 performs a correction calculation of the heat transfer coefficient according to the procedure shown in FIG. Ru. That is, first, in step 141, the metal temperature distribution T〓 ^/1 (j) obtained by the first metal temperature distribution calculation section 12 is calculated (where i=0, 1, ..., N).
and the metal temperature distribution T〓 ^/1 (j) obtained by the second metal temperature distribution calculation unit 10, (where i=0, 1, ...,
N), the sum-of-squares evaluation function F of the deviation is determined by the following equation (17).

F= _N 〓 ⁱ⁼⁰ ...(17) Next, in step 142, the function F is compared with the set tolerance value ε, and the metal temperature distribution obtained by the first metal temperature distribution calculation section 12 is calculated. T〓 ^/1 (j)
Determine whether or not it is correct. Note that the allowable value ε regarding the temperature distribution difference is set based on whether the corresponding stress value is allowable in stress evaluation or the like. Then, as a result of the size comparison, the tolerance value ε
When it is equal to or less than , the process proceeds to step 148, and the heat transfer coefficient correction process is ended without correcting the value of the heat transfer coefficient h. Thereby, the thermal stress calculation unit 17 executes calculations described below based on the metal temperature distribution output from the first metal temperature distribution calculation unit 12. On the other hand, if it is larger than the allowable value ε, in step 143, the amount of heat input from the internal fluid to the metal at the inner surface of the metal (r=a) is calculated as the temperature distribution T〓 ^/1 (j), T〓 ^/1 Based on (j), Q ₁ and Q ₂ are calculated using the following equations (18) and (19), respectively.

Q ₁ = −λ∂T/∂r｜ _rna = h(T〓 ^/1 (j)−T〓 ^/0 ) ……(18) Q ₂ = −λ∂T〓/∂r｜ _rna ……(19 ) Then, in step 144, the heat input amounts Q ₁ and Q ₂ are compared. When Q ₁ is equal to or greater than Q ₂ , it is an overestimation of the heat input in the first method, so
In step 145, the heat transfer coefficient is reduced by a predetermined minute amount Δh. If Q ₁ is smaller than Q ₂ for the reaction, the heat input is insufficiently evaluated, and therefore, in step 146, the heat transfer coefficient h is corrected by increasing it by Δh.

Next, in step 147, using the corrected heat transfer coefficient h, a metal temperature distribution calculation is performed on a trial basis using the same method as in the first metal temperature distribution calculation section 12, and the process returns to step 141. Until the sum of squares evaluation function F becomes less than or equal to the tolerance value ε,
While modifying the heat transfer coefficient h in increments of Δh, the processes from steps 141 to 147 are repeatedly executed to obtain the optimum heat transfer coefficient h.

Note that the heat transfer coefficient h may be corrected by the procedure shown in FIG. That is, in the embodiment shown in FIG. 8, the metal temperature distribution calculation is not attempted each time the correction is made in Δh increments, but the heat transfer coefficient h is corrected in Δh increments only once for each entire monitoring calculation cycle of thermal stress monitoring. It was designed to do this.

Next, the thermal stress calculation method in the thermal stress calculation section 17 will be explained.

The thermal stress at time t at a point with radius r from the center of the cylinder is expressed in polar coordinates using the following equation: radial thermal stress σ _r (r, t), circumferential thermal stress σ 〓 (r, t), axial thermal stress σ When _z (r, t), these are calculated using the following formula.

σ _r (r,t)=Eβ/1−ν{1/b ² −a ² (1−a ² /r ²
)∫ ^b/a T(r,t)rdr−1/r ² ∫ ^r/a T(r,t)rdr
}...(11) σ〓(r,t)=Eβ/1−ν{1/b ² −a ² (1+a ² /r
² )∫ ^b/a T(r,t)rdr−1/r ² ∫ ^r/a T(r,t)rd
r−T(,t)}
...(12) σ _z (r, t)=Eβ/1−ν{2/b ² −a ² ∫ ^b/a T(r
, t)rdr−T(r,t)}...(13) where, E: Young's modulus β: coefficient of linear expansion ν: Boisson's ratio (=constant) Young's modulus E and coefficient of linear expansion β are determined by metal temperature distribution calculation The metal volume average temperature is determined based on the temperature, and this is selected as a parameter from a constant table stored in advance.

The thermal stress is most severe at the inner surface of the metal, that is, at the point r = a (corresponding to the dividing point 0 in Figure 6).
Calculate from the following formula where = a.

σ _r (a, t)=0 ...(14) σ〓(a, t)=Eβ/1−ν{2/b ² −a ² ∫ ^b
/a T(r,t)rdr−T(a,t)}……(15) σ _z (a,t)=Eβ/1−ν{2/b ² −a ² ∫ ^b/
a T(r,t)rdr−T(a,t)}……(16) Here, σ _r (a, t)=0, σ〓(a, t)=σ _z (a
，
Since the relationship t) holds true, the thermal stress in the general part inside the metal cylinder is evaluated using σ〓(a, t) as a representative value.

Then, based on this representative value of thermal stress, the lifetime consumption of the evaluation point is calculated according to the method disclosed in Japanese Patent Application No. 57-223939, as described above.

〔Effect of the invention〕

As explained above, according to the present invention, thermal stress can be determined with high precision and reliability is improved, so thermal stress monitoring and lifetime consumption management of boilers can be performed with high precision and high reliability. be able to.

[Brief explanation of the drawing]

Fig. 1 is an overall configuration diagram of an embodiment of the device to which the present invention is applied, Fig. 2 is a detailed functional block diagram of the main parts of the embodiment shown in Fig. 1, and Figs. 3 to 6 show metal temperature distribution calculations. The explanatory diagrams, FIG. 7 and FIG. 8, are flowcharts showing the processing procedure for modifying the heat transfer coefficient, respectively. 10...Second metal temperature distribution calculation section, 11...
Starting circuit, 12...first metal temperature distribution calculation section,
13...Heat transfer coefficient calculation unit, 14...Heat transfer coefficient correction unit, 15...Diagnostic circuit, 16...Correction lock circuit, 17...Thermal stress calculation unit, 18...Life consumption calculation unit.

Claims (1)

[Claims] 1. Detect the state quantity of the internal fluid at the stress monitoring evaluation point set on the boiler member, estimate the heat transfer coefficient of the interface between the internal fluid and the member at the evaluation point based on this detected data, and a first method of determining the temperature distribution in the thickness direction of the member by determining the amount of heat input to the member based on the estimated transmissivity value and the detected data; detecting the inner surface temperature and outer surface temperature of the member at the evaluation point; There is a second method for determining the amount of heat input to the member and the temperature distribution in the thickness direction of the member based on the inner surface temperature and outer surface temperature; the temperature distribution obtained by the first method and the second method is compared. ,
When the difference is greater than or equal to the allowable value set based on the standard of whether or not the difference in the corresponding stress values is allowable, the heat input of each of the above methods corresponding to each temperature distribution is compared, and the heat input of the first method is compared. When the heat input is larger than the heat input of the second method, the estimated value of the heat transfer coefficient of the first method is reduced, and when it is small, the estimated value of the heat transfer coefficient is increased, and the corrected heat transfer coefficient is used. A boiler thermal stress monitoring method, characterized in that the thermal stress at the evaluation point is calculated based on the temperature distribution obtained by the first method.