JP7728686B2 - Denitration control device, denitration control method and program - Google Patents

Description

The present invention relates to a denitration control device, a denitration control method, and a program.

The exhaust gas from a gas turbine contains nitrogen oxides (NOx), an environmental pollutant. A denitration system supplies ammonia (NH ₃ ) as a reducing agent to the exhaust gas, causing the nitrogen oxides contained in the exhaust gas to react with the ammonia, breaking them down into nitrogen and water, thereby reducing the amount of nitrogen oxides emitted from the chimney and managing environmental regulation values.

Patent Document 1, paragraph 0005, discloses a method for controlling the amount of ammonia injected into a denitration system. This method calculates a basic ammonia injection amount corresponding to the amount of nitrogen oxides calculated from the deviation between the set value of the nitrogen oxide concentration at the denitration system outlet or chimney inlet and the measured value of the nitrogen oxide concentration at the denitration system inlet, and corrects this by feeding back the nitrogen oxide concentration at the denitration system outlet. The method then sets a preliminary ammonia injection amount in accordance with fluctuations in the gas turbine load and makes further corrections.

Japanese Patent Application Laid-Open No. 2003-80026

However, analyzers that measure the nitrogen oxide concentration at the inlet of a denitration device have delays in the time it takes to take in sampled gas and the time it takes to analyze it. For example, even if the concentration of nitrogen oxides contained in exhaust gas passing through a denitration device is measured at the inlet of the denitration device and the amount of ammonia to be supplied to that exhaust gas (the exhaust gas whose nitrogen oxide concentration has been measured) is calculated, by the time the calculated amount of ammonia is supplied to the denitration device, the exhaust gas (the exhaust gas whose nitrogen oxide concentration has been measured) has already passed through the denitration device.
In other words, when an analyzer is used to measure the nitrogen oxide concentration at the inlet of the denitration device, the exhaust gas whose nitrogen oxide concentration is measured by the analyzer as it passes through the denitration device (the exhaust gas whose nitrogen oxide concentration is to be measured) will not match with the exhaust gas to which ammonia is supplied in an amount calculated based on the nitrogen oxide concentration measured by the analyzer (the exhaust gas to which ammonia is supplied), resulting in an excess or deficiency in the amount of ammonia supplied, and as a result, denitration control cannot be performed appropriately.

In the technology described in Example 2 of Patent Document 1, in order to correct an excess or deficiency in the amount of ammonia supplied due to an analysis delay in the analyzer (inlet nitrogen oxide detection means), corrections (advance correction, feedback correction, nitrogen dioxide compensation amount correction) are made to the amount of ammonia supplied (basic ammonia amount) set by the nitrogen oxide concentration prediction circuit (inlet nitrogen oxide concentration estimation means) at the inlet of the denitrification device, such as a correction (advance correction) that proactively adds the amount of ammonia in response to changes in the gas turbine load. However, there are cases in which an excess or deficiency in the amount of ammonia supplied due to inconsistencies in the timing of ammonia supply caused by a delay in the analysis of nitrogen oxide concentration is not resolved.

In addition, maintaining the measurement accuracy of analyzers that measure nitrogen oxide concentrations requires regular maintenance such as calibration and part replacement, resulting in high running costs.

In light of the above, the present invention aims to provide a denitration control device, method, and program that can perform denitration control by appropriately controlling the amount of ammonia supplied without installing an analyzer that measures the concentration of nitrogen oxides at the inlet of the denitration device.

One aspect of the present invention is a denitration control device that controls the amount of ammonia supplied to a denitration device that reduces the nitrogen oxide content in exhaust gas containing nitrogen oxides generated by combustion in a gas turbine. The denitration control device does not have an analyzer installed at the inlet of the denitration device to measure the concentration of nitrogen oxides, and is equipped with a nitrogen oxide concentration prediction model that predicts the concentration of nitrogen oxides at the inlet of the denitration device through model calculations.

In one aspect of the denitrification control device of the present invention, the combustor of the gas turbine performs premixed combustion and diffusion combustion, and the gas turbine is equipped with a combustor bypass valve that controls the fuel/air ratio, which is the ratio of fuel to air used for combustion in the combustor. The nitrogen oxide concentration prediction model may include: a first three-dimensional function unit that calculates the nitrogen oxide concentration prediction value based on a first fuel amount that is the fuel amount used for the premixed combustion, a second fuel amount that is the fuel amount used for the diffusion combustion, and a predetermined first relationship between the first fuel amount, the second fuel amount, and a nitrogen oxide concentration prediction value that is a predicted value of the nitrogen oxide concentration; and a multiplier that corrects the nitrogen oxide concentration prediction value calculated by the first three-dimensional function unit based on a first nitrogen oxide concentration correction amount calculated according to the fuel/air ratio and an opening of the combustor bypass valve.

In one embodiment of the denitration control device of the present invention, the nitrogen oxide concentration prediction model may include a second three-dimensional function unit that calculates the second nitrogen oxide concentration correction amount based on a second predetermined relationship between the combustion air temperature, which is the temperature of the air used for combustion in the combustor, the combustion air humidity, which is the humidity of the air used for combustion in the combustor, and a second nitrogen oxide concentration correction amount that further corrects the nitrogen oxide concentration prediction value corrected by the multiplier, and the combustion air temperature and combustion air humidity.

In one embodiment of the present invention, the nitrogen oxide concentration prediction model includes a process signal selection unit that selects a plurality of process signals related to combustion in the combustor of the gas turbine, the process signals being assumed to affect the concentration of nitrogen oxides at the inlet of the denitration device, and a process signal selection unit that performs a multiple regression analysis using the plurality of process signals selected by the process signal selection unit as explanatory variables and the concentration of nitrogen oxides as a target variable, and the process signal selection unit selects a value used to predict the concentration of nitrogen oxides from the plurality of process signals selected by the process signal selection unit based on a t-value of the multiple regression analysis indicating the degree of influence of the plurality of process signals selected by the process signal selection unit on the concentration of nitrogen oxides. The nitrogen oxide concentration prediction model may include a compatibility evaluation unit that evaluates the compatibility of the plurality of selected process signals selected by the process signal selection unit based on a correction R2 of multiple regression analysis, the compatibility evaluation unit determines a plurality of regression coefficients in a multiple regression equation by which the plurality of selected process signals selected by the process signal selection unit are respectively multiplied, and the nitrogen oxide concentration prediction model may include a nitrogen oxide concentration prediction unit that predicts the concentration of nitrogen oxides at the inlet of the denitrification device based on the plurality of selected process signals selected by the process signal selection unit, the plurality of regression coefficients determined by the compatibility evaluation unit, and an intercept of the multiple regression equation.

In one aspect of the denitration control device of the present invention, the combustor of the gas turbine performs premixed combustion and diffusion combustion, the gas turbine is equipped with a combustor bypass valve that controls the fuel/air ratio, which is the ratio of fuel to air used for combustion in the combustor, and the compressor of the gas turbine is equipped with an IGV (inlet guide vane), and the multiple process signals selected by the process signal selection unit may include multiple process signals from among a process signal indicating a combustion fuel flow rate, a process signal indicating a combustion air flow rate, a process signal indicating a diffusion combustion fuel distribution ratio, a process signal indicating a premixed combustion fuel distribution ratio, a process signal indicating an IGV angle, a process signal indicating a combustor bypass valve opening, a process signal indicating a gas turbine exhaust gas temperature, a process signal indicating a gas turbine blade path temperature, a process signal indicating an air compressor inlet temperature, a process signal indicating an air compressor outlet temperature, a process signal indicating an air compressor inlet pressure, and a process signal indicating combustion air humidity.

In one aspect of the denitration control device of the present invention, the compressor of the gas turbine is equipped with an inlet guide vane (IGV), and the nitrogen oxide concentration prediction model includes a two-dimensional function calculator that calculates a predicted value of the reaction rate of the nitrogen oxide concentration based on the fuel flow rate used for combustion in the combustor of the gas turbine and a predetermined relationship between the fuel flow rate and a predicted value of the reaction rate of the nitrogen oxide concentration at the inlet of the denitration device. The nitrogen oxide concentration at the inlet of the denitration device may be predicted by correcting the predicted value of the reaction rate of the nitrogen oxide concentration calculated by the two-dimensional function calculator based on the combustion air volume, which is the amount of air used for combustion in the combustor, and the IGV angle, which is the angle of the IGV.

In one aspect of the denitration control device of the present invention, the combustor of the gas turbine performs premixed combustion and diffusion combustion, the gas turbine is equipped with a combustor bypass valve that controls the fuel/air ratio, which is the ratio of fuel to air used for combustion in the combustor, and the compressor of the gas turbine is equipped with an IGV (inlet guide vane), and the nitrogen oxide concentration prediction model includes a first three-dimensional function unit that calculates the nitrogen oxide concentration prediction value based on a first fuel amount that is the fuel amount used in the premixed combustion, a second fuel amount that is the fuel amount used in the diffusion combustion, and a predetermined first relationship between the first fuel amount, the second fuel amount, and a nitrogen oxide concentration prediction value that is a predicted value of the nitrogen oxide concentration, and a first nitrogen oxide concentration correction amount calculated according to the fuel/air ratio and an opening degree of the combustor bypass valve. a combustion air temperature that is the temperature of the air used for combustion in the combustor, a combustion air humidity that is the humidity of the air used for combustion in the combustor, and a second nitrogen oxide concentration correction amount that further corrects the nitrogen oxide concentration predicted value corrected by the multiplier; a three-dimensional function model including a second three-dimensional function unit that calculates the second nitrogen oxide concentration correction amount based on a second predetermined second relationship between the combustion air temperature, the combustion air humidity, and a second nitrogen oxide concentration correction amount that further corrects the nitrogen oxide concentration predicted value corrected by the multiplier; a process signal selection unit that selects a plurality of process signals related to combustion in the combustor, the plurality of process signals being assumed to affect the nitrogen oxide concentration at the inlet of the denitration device; and a multiple regression analysis that uses the plurality of process signals selected by the process signal selection unit as explanatory variables and the nitrogen oxide concentration as a response variable, and a process signal selection unit that selects a plurality of selected process signals from the plurality of process signals selected by the process signal selection unit based on a t value of multiple regression analysis that indicates the degree of influence of each process signal on the concentration of nitrogen oxides; a compatibility evaluation unit that evaluates the compatibility of the plurality of selected process signals selected by the process signal selection unit based on a correction R2 of the multiple regression analysis and determines a plurality of regression coefficients to be multiplied by the plurality of selected process signals selected by the process signal selection unit in a multiple regression equation; and a nitrogen oxide concentration prediction unit that predicts the concentration of nitrogen oxides at an inlet of the denitration device based on the plurality of selected process signals selected by the process signal selection unit, the plurality of regression coefficients determined by the compatibility evaluation unit, and an intercept of the multiple regression equation. The system includes a two-dimensional function calculator that calculates a predicted value of the reaction rate of the nitrogen oxide concentration based on a multiple regression analysis model and a reaction rate equation that is a predetermined relationship between the fuel flow rate used for combustion in the combustor and a predicted value of the reaction rate of the nitrogen oxide concentration at the inlet of the denitration device, and the predicted value of the reaction rate of the nitrogen oxide concentration calculated by the two-dimensional function calculator is corrected based on the combustion air volume, which is the amount of air used for combustion in the combustor, and the IGV angle, which is the angle of the IGV, to thereby provide at least two reaction rate equation models that predict the nitrogen oxide concentration at the inlet of the denitration device, and one of the at least two models may be selected and used, or a combination of the at least two models may be used, depending on the characteristics of the gas turbine plant to which the denitration control device is applied.

One aspect of the present invention is a denitration control method for a denitration control device that includes an ammonia supply amount control step that controls the amount of ammonia supplied to a denitration device that reduces the nitrogen oxide content in exhaust gas containing nitrogen oxides generated by combustion in a gas turbine. The denitration control method does not include an analyzer that measures the concentration of nitrogen oxides at the inlet of the denitration device, and includes a nitrogen oxide concentration prediction step that predicts the concentration of nitrogen oxides at the inlet of the denitration device using a model calculation.

One aspect of the present invention is a program for causing a computer installed in a denitration control device to execute an ammonia supply amount control step for controlling the amount of ammonia supplied to a denitration device to reduce the nitrogen oxide content in exhaust gas containing nitrogen oxides generated by combustion in a gas turbine, and a nitrogen oxide concentration prediction step for predicting the concentration of nitrogen oxides at the inlet of the denitration device through model calculations, wherein no analyzer for measuring the concentration of nitrogen oxides is installed at the inlet of the denitration device.

The present invention provides a denitration control device, method, and program that can perform denitration control by appropriately controlling the amount of ammonia supplied without installing an analyzer that measures the concentration of nitrogen oxides at the inlet of the denitration device.

1 is a diagram illustrating an example of a gas turbine plant (gas turbine power generation facility) GP to which a denitration control device 1 of a first embodiment is applied. FIG. 2 is a diagram showing an example of a nitrogen oxide concentration prediction model 11 of the denitration control device 1 of the first embodiment. FIG. 10 is a diagram illustrating an example of a first relation used in a three-dimensional function unit A10. FIG. 10 is a diagram illustrating an example of a first relation used in a three-dimensional function unit A10. FIG. 2 is a diagram for explaining an example of processing executed in the denitration control device 1 of the first embodiment. FIG. 10 is a diagram showing an example of a nitrogen oxide concentration prediction model 11 of the denitration control device 1 of the second embodiment. FIG. 11 is a diagram showing an example of a nitrogen oxide concentration prediction model 11 of a denitration control device 1 of a third embodiment.

Embodiments of the denitration control device, denitration control method, and program of the present invention will be described below with reference to the drawings.

[First embodiment]
FIG. 1 is a diagram showing an example of a gas turbine plant (gas turbine power generation facility) GP to which a denitration control device 1 according to the first embodiment is applied.
The gas turbine plant GP to which the denitration control device 1 of the first embodiment is applied is equipped with the denitration control device 1, an ammonia flow control valve 3, a chimney inlet nitrogen oxide concentration analyzer 4, a gas turbine 5, a flue 6, a denitration device 7, and a chimney 8.
The denitration control device 1 controls the amount of ammonia supplied to the denitration device 7. The ammonia flow rate control valve 3 adjusts the flow rate of ammonia supplied to the denitration device 7. The stack inlet nitrogen oxide concentration analyzer 4 measures the concentration of nitrogen oxides in the exhaust gas at the inlet of the stack 8.

The gas turbine 5 has the same functions as, for example, the gas turbine described in Japanese Patent No. 4831820, and includes a compressor, a combustor, a turbine, and a combustor bypass valve. The combustor of the gas turbine 5 performs premixed combustion and diffusion combustion. The combustor bypass valve controls the fuel/air ratio, which is the ratio of fuel to air used for combustion in the combustor. The compressor includes inlet guide vanes (IGVs). The IGVs adjust the amount of air used for combustion in the combustor of the gas turbine 5 by changing their angle (opening). The flue 6 connects the gas turbine 5 to a chimney 8.
The denitration device 7 is disposed in the flue 6. The denitration device 7 supplies ammonia to the exhaust gas containing nitrogen oxides generated by combustion in the gas turbine 5, thereby reducing the content of nitrogen oxides in the exhaust gas.

In the example shown in Figure 1, exhaust gas emitted by combustion in a gas turbine 5 passes through a flue 6 and is released into the outside air from a chimney 8. During this process, nitrogen oxides contained in the exhaust gas react with ammonia in a denitration device 7, which has a catalyst that uses ammonia as a reducing agent, and are decomposed into nitrogen and water. The exhaust gas, whose nitrogen oxide content has been reduced in the denitration device 7, is released into the outside air from a chimney 8.

In accordance with the Environmental Impact Assessment Act, the travel time average value of unreacted nitrogen oxides contained in the exhaust gas released into the outside air from the chimney 8 is required to be controlled to below a regulated value, which is determined by relevant laws and regulations and agreements with the government.
The amount of ammonia used as a reducing agent in the denitration device 7 to reduce nitrogen oxides in the exhaust gas from the gas turbine 5 is calculated by the denitration control device 1. The amount of ammonia calculated by the denitration control device 1 is supplied to the denitration device 7 by operating the ammonia flow rate control valve 3.

In a typical gas turbine power generation facility, in order to calculate the amount of ammonia in the denitration control device 1, an analyzer 2 (see FIG. 1) that measures the nitrogen oxide concentration in the exhaust gas generated by combustion in the gas turbine 5 is installed together with a stack inlet nitrogen oxide concentration analyzer 4 that analyzes the nitrogen oxide concentration in the exhaust gas released into the outside air from a stack 8.
On the other hand, in the gas turbine plant GP to which the denitration control device 1 of the first embodiment is applied, the analyzer 2 (see FIG. 1) for measuring the concentration of nitrogen oxides is not installed at the inlet of the denitration device 7 .

In the example shown in FIG. 1, the denitration control device 1 includes a nitrogen oxide concentration prediction model 11, a nitrogen oxide treatment amount calculation unit 12, an ammonia flow rate calculation unit 13, and a feedback correction amount calculation unit 14.
The nitrogen oxide concentration prediction model 11 predicts, by model calculation, the concentration of nitrogen oxides in the exhaust gas at the inlet of the denitration device 7. The nitrogen oxide treatment amount calculation unit 12 calculates the amount of nitrogen oxides that needs to be treated (decomposed into nitrogen and water) in the denitration device 7 (nitrogen oxide treatment amount) based on the nitrogen oxide concentration in the exhaust gas predicted by the nitrogen oxide concentration prediction model 11.
Ammonia flow rate calculation unit 13 calculates the amount of ammonia required to treat the amount of nitrogen oxides calculated by nitrogen oxide treatment amount calculation unit 12, and generates and outputs a control signal for ammonia flow rate control valve 3. Feedback correction amount calculation unit 14 calculates a feedback correction amount according to the deviation between the nitrogen oxide concentration in the exhaust gas at the inlet of chimney 8 measured by chimney inlet nitrogen oxide concentration analyzer 4 and a preset control value for the nitrogen oxide concentration in the exhaust gas discharged from chimney 8, and outputs the feedback correction amount to ammonia flow rate calculation unit 13.

In other words, the nitrogen oxide treatment amount calculation unit 12 of the denitrification control device 1 calculates the nitrogen oxide treatment amount based on the nitrogen oxide concentration predicted by the nitrogen oxide concentration prediction model 11. The ammonia flow rate calculation unit 13 calculates the amount of ammonia required to treat the nitrogen oxides and outputs a control amount command to the ammonia flow rate control valve 3. In addition, the feedback correction amount calculation unit 14 provides the ammonia flow rate calculation unit 13 with a feedback correction amount corresponding to the deviation between the analysis value of the stack inlet nitrogen oxide concentration analyzer 4 and the control value of the nitrogen oxide concentration in the exhaust gas discharged from the stack 8.

FIG. 2 is a diagram showing an example of the nitrogen oxide concentration prediction model 11 of the denitration control device 1 of the first embodiment.
2, the nitrogen oxide concentration prediction model 11 is configured by a three-dimensional function model 11A. The three-dimensional function model 11A includes a three-dimensional function unit A10, a three-dimensional function unit A11, a multiplier A12, a subtractor A13, a combustor bypass valve opening information acquisition unit A14, a fuel/air ratio calculation unit A15, a two-dimensional function unit A16, a fixed value setting unit A17, and a switch A18.
The three-dimensional function calculator A10 calculates a nitrogen oxide concentration predicted value, which is a predicted value of the concentration of nitrogen oxides in the exhaust gas at the inlet of the denitration device 7. In detail, the three-dimensional function calculator A10 calculates the nitrogen oxide concentration predicted value based on a first fuel amount which is the fuel amount used for premixed combustion in the combustor of the gas turbine 5, a second fuel amount which is the fuel amount used for diffusion combustion in the combustor of the gas turbine 5, and a predetermined first relationship (see FIGS. 3 and 4 ). The first relationship is a relationship between the first fuel amount, the second fuel amount, and the nitrogen oxide concentration predicted value.

3 and 4 are diagrams showing an example of the first relationship used in the three-dimensional function unit A 10. In detail, Fig. 3 shows the first relationship in the form of a matrix table, and Fig. 4 shows the first relationship three-dimensionally (stereoscopically).
3, the horizontal axis indicates the amount of the first fuel used for premixed combustion in the combustor of the gas turbine 5, and the vertical axis indicates the amount of the second fuel used for diffusion combustion in the combustor of the gas turbine 5. "X0", "X1", "X2", "X3", "X4", ..., "Xi" on the horizontal axis of Fig. 3 indicate values of the first fuel amount, "Y0", "Y1", "Y2", "Y3", "Y4", ..., "Yi" on the vertical axis of Fig. 3 indicate values of the second fuel amount, and "n00", ..., "n11", ..., "n22", ..., "n33", ..., "n44", ..., "nii" indicate nitrogen oxide concentration predicted values calculated by the three-dimensional function unit A10.
For example, when the value of the first fuel amount is "X0" and the value of the second fuel amount is "Y0", the three-dimensional function unit A10 calculates a nitrogen oxide concentration predicted value "n00". When the value of the first fuel amount is "X1" and the value of the second fuel amount is "Y2", the three-dimensional function unit A10 calculates a nitrogen oxide concentration predicted value "n12". When the value of the first fuel amount is "X4" and the value of the second fuel amount is "Yi", the three-dimensional function unit A10 calculates a nitrogen oxide concentration predicted value "n4i".
The relationship (first relationship) between the first fuel amount values "X0", ..., "Xi", the second fuel amount values "Y0", ..., "Yi", and the nitrogen oxide concentration predicted values "n00", ..., "nii" shown in FIG. 3 is set in advance by, for example, conducting an experiment or a simulation.

In FIG. 4, the X-axis represents the amount of the first fuel used for premixed combustion in the combustor of the gas turbine 5, the Y-axis represents the amount of the second fuel used for diffusion combustion in the combustor of the gas turbine 5, and the Z-axis represents the predicted value of the nitrogen oxide concentration calculated by the three-dimensional function calculator A10.
In the example shown in FIG. 4, the predicted nitrogen oxide concentration value increases as the first fuel amount increases, and the predicted nitrogen oxide concentration value increases as the second fuel amount increases.
In FIG. 4 , the relationship (first relationship) between the first fuel amount indicated on the X-axis, the second fuel amount indicated on the Y-axis, and the predicted nitrogen oxide concentration value indicated on the Z-axis is set in advance by, for example, conducting an experiment or a simulation.

In the example shown in FIG. 2, a fuel/air ratio calculation unit A15 calculates the ratio of fuel to air used for combustion in the combustor of the gas turbine 5 (fuel/air ratio).
The two-dimensional function unit A16 calculates a first nitrogen oxide concentration correction amount for correcting the predicted value of nitrogen oxide concentration calculated by the three-dimensional function unit A10, based on the fuel/air ratio calculated by the fuel/air ratio calculation unit A15. More specifically, the two-dimensional function unit A16 calculates the first nitrogen oxide concentration correction amount based on the fuel/air ratio calculated by the fuel/air ratio calculation unit A15 and a preset relationship between the fuel/air ratio and the first nitrogen oxide concentration correction amount. The relationship between the fuel/air ratio and the first nitrogen oxide concentration correction amount is set in advance, for example, by conducting experiments, simulations, etc.

The combustor bypass valve opening information acquisition unit A14 acquires information indicating the opening of the combustor bypass valve that controls the fuel/air ratio.
The fixed value setter A17 outputs the value "1" as the first nitrogen oxide concentration correction amount.
The switch A18 switches the value of the first nitrogen oxide concentration correction amount output to the multiplier A12. Specifically, when the combustor bypass valve is open, the switch A18 outputs the value "1" output from the fixed value setter A17 to the multiplier A12 as the first nitrogen oxide concentration correction amount. In this case, the nitrogen oxide concentration predicted value calculated by the three-dimensional function unit A10 is not corrected by the multiplier A12. On the other hand, when the combustor bypass valve is not open, the switch A18 outputs the first nitrogen oxide concentration correction amount calculated by the two-dimensional function unit A16 to the multiplier A12. In this case, the nitrogen oxide concentration predicted value calculated by the three-dimensional function unit A10 is corrected in the multiplier A12 based on the first nitrogen oxide concentration correction amount calculated by the two-dimensional function unit A16.

The multiplier A12 corrects the predicted value of the nitrogen oxide concentration calculated by the three-dimensional function calculator A10 by multiplying the predicted value of the nitrogen oxide concentration calculated by the three-dimensional function calculator A10 by the value (first nitrogen oxide concentration correction amount) output from the switch A18.
That is, the multiplier A12 corrects the predicted nitrogen oxide concentration value calculated by the three-dimensional function calculator A10 based on the first nitrogen oxide concentration correction amount calculated by the two-dimensional function calculator A16 in accordance with the fuel/air ratio and the opening of the combustor bypass valve.

The three-dimensional function calculator A11 calculates a second nitrogen oxide concentration correction amount for further correcting the nitrogen oxide concentration predicted value corrected by the multiplier A12, based on the combustion air temperature, which is the temperature of the air used for combustion in the combustor of the gas turbine 5, and the combustion air humidity, which is the humidity of the air used for combustion in the combustor of the gas turbine 5.
Specifically, the three-dimensional function calculator A11 calculates the second nitrogen oxide concentration correction amount based on the combustion air temperature obtained by, for example, a temperature sensor, the combustion air humidity obtained by, for example, a humidity sensor, and a predetermined second relationship. The second relationship, which is the relationship between the combustion air temperature, the combustion air humidity, and the second nitrogen oxide concentration correction amount, is set in advance by, for example, conducting experiments or simulations.

The subtractor A13 further corrects the nitrogen oxide concentration predicted value corrected by the multiplier A12 by subtracting the second nitrogen oxide concentration correction amount calculated by the three-dimensional function unit A11 from the nitrogen oxide concentration predicted value corrected by the multiplier A12.
In the example shown in FIG. 2 , the nitrogen oxide concentration prediction model 11 (three-dimensional function model 11A) outputs the value output from the subtractor A13 as a predicted value of the nitrogen oxide concentration in the exhaust gas at the inlet of the denitration device 7.

In other words, the denitration control device 1 of the first embodiment takes into account that the concentration of nitrogen oxides generated by combustion in the combustor of the gas turbine 5 depends on the fuel distribution ratio between diffusion combustion and premixed combustion in the combustor of the gas turbine 5. In detail, the denitration control device 1 of the first embodiment sets predicted data (predicted nitrogen oxide concentration values) corresponding to the analysis values of the analyzer 2 (see FIG. 1) at the inlet of the denitration device 7 in a matrix table (see FIG. 3) with the respective fuel distribution ratios on the X axis (amount of fuel used in premixed combustion) and the Y axis (amount of fuel used in diffusion combustion). This allows the three-dimensional function calculator A10 to calculate (predict) the predicted nitrogen oxide concentration values corresponding to the analysis values of the analyzer 2, without the need to install the analyzer 2 (see FIG. 1) at the inlet of the denitration device 7.

The denitration control device 1 of the first embodiment also takes into account the fact that the concentration of nitrogen oxides generated by combustion in the combustor of the gas turbine 5 also depends on the ratio of fuel to air (fuel/air ratio) used for combustion in the combustor. In particular, in the denitration control device 1 of the first embodiment, the fuel/air ratio calculation unit A15 calculates the ratio of fuel to air (fuel/air ratio) used for combustion in the combustor of the gas turbine 5, and the two-dimensional function unit A16 calculates a first nitrogen oxide concentration correction amount for correcting the predicted nitrogen oxide concentration value calculated by the three-dimensional function unit A10 based on the fuel/air ratio calculated by the fuel/air ratio calculation unit A15.

Furthermore, the denitration control device 1 of the first embodiment takes into account that the concentration of nitrogen oxides generated by combustion in the combustor of the gas turbine 5 also depends on the temperature and humidity of the air used for combustion in the combustor. In detail, the denitration control device 1 of the first embodiment sets the second nitrogen oxide concentration correction amount in a matrix table (not shown) similar to the matrix table shown in FIG. 3, in which the temperature and humidity of the air used for combustion in the combustor are the X-axis and Y-axis, respectively, and the three-dimensional function calculator A11 calculates the second nitrogen oxide concentration correction amount from the temperature and humidity of the air used for combustion in the combustor.

As mentioned above, when analyzer 2 (see Figure 1) is used to measure the nitrogen oxide concentration at the inlet of the denitration device, there is a delay in the analysis time (typically about one minute). This means that the exhaust gas whose nitrogen oxide concentration is measured by analyzer 2 as it passes through denitration device 7 does not match the exhaust gas to which the amount of ammonia calculated based on the nitrogen oxide concentration measured by analyzer 2 is supplied. This results in an excess or deficiency in the amount of ammonia supplied, and as a result, denitration control cannot be performed appropriately.

In contrast, in a gas turbine plant GP to which the denitration control device 1 of the first embodiment is applied, the analyzer 2 (see Figure 1) that measures the nitrogen oxide concentration is not installed at the inlet of the denitration device 7, thereby eliminating the problem of analysis time delays. Furthermore, in the denitration control device 1 of the first embodiment, the ammonia supply amount is controlled based on the nitrogen oxide concentration predicted by the nitrogen oxide concentration prediction model 11 (i.e., the nitrogen oxide concentration without analysis time delay), allowing for appropriate denitration control.

FIG. 5 is a diagram for explaining an example of processing executed in the denitration control device 1 of the first embodiment.
In the example shown in FIG. 5, in step S1, the nitrogen oxide concentration prediction model 11 (three-dimensional function model 11A) of the denitration control device 1 predicts the concentration of nitrogen oxides in the exhaust gas at the inlet of the denitration device 7.
Next, in step S2, the ammonia flow rate calculation unit 13 of the denitration control device 1 calculates the amount of ammonia to be supplied that is necessary to reduce the content of nitrogen oxides in the exhaust gas, and generates and outputs a control signal to the ammonia flow rate control valve 3. As a result, the ammonia flow rate control valve 3 is driven, and the amount of ammonia calculated by the ammonia flow rate calculation unit 13 is supplied to the denitration device 7. That is, in the denitration device 7, the amount of ammonia to be supplied to the exhaust gas that contains nitrogen oxides is controlled.

Second Embodiment
A second embodiment of the denitration control device, the denitration control method, and the program according to the present invention will now be described.
The denitration control device 1 of the second embodiment is configured similarly to the denitration control device 1 of the first embodiment described above, except for the points described below. Therefore, the denitration control device 1 of the second embodiment can achieve the same effects as the denitration control device 1 of the first embodiment described above, except for the points described below.

The gas turbine plant GP to which the denitration control device 1 of the second embodiment is applied has the same configuration as the gas turbine plant GP to which the denitration control device 1 of the first embodiment shown in FIG. 1 is applied.
That is, the gas turbine plant GP to which the denitration control device 1 of the second embodiment is applied is equipped with the denitration control device 1, an ammonia flow control valve 3, a chimney inlet nitrogen oxide concentration analyzer 4, a gas turbine 5, a flue 6, a denitration device 7, and a chimney 8, and the analyzer 2 (see FIG. 1 ) that measures the concentration of nitrogen oxides is not installed at the inlet of the denitration device 7.
The denitration control device 1 of the second embodiment controls the amount of ammonia supplied to the denitration device 7, similarly to the denitration control device 1 of the first embodiment.

In one example of a gas turbine plant GP to which the denitration control device 1 of the second embodiment is applied, similar to the example shown in FIG. 1 , the denitration control device 1 includes a nitrogen oxide concentration prediction model 11, a nitrogen oxide treatment amount calculation unit 12, an ammonia flow rate calculation unit 13, and a feedback correction amount calculation unit 14.
The nitrogen oxide concentration prediction model 11 of the denitration control device 1 of the second embodiment predicts the concentration of nitrogen oxides in the exhaust gas at the inlet of the denitration device 7 by model calculation, similar to the nitrogen oxide concentration prediction model 11 of the denitration control device 1 of the first embodiment.

FIG. 6 is a diagram showing an example of the nitrogen oxide concentration prediction model 11 of the denitration control device 1 of the second embodiment.
6, the nitrogen oxide concentration prediction model 11 is configured by a multiple regression analysis model 11B. The multiple regression analysis model 11B includes a process signal selection unit B20, a process signal classification unit B21, a compatibility evaluation unit B22, and a nitrogen oxide concentration prediction unit B23.
The process signal selection unit B20 selects a plurality of process signals P1, P2, P3, P4, P5, ..., Pi that are related to combustion in the combustor of the gas turbine 5 and are assumed to affect the concentration of nitrogen oxides at the inlet of the denitration device 7.
The multiple process signals P1, P2, P3, P4, P5, ..., Pi selected by the process signal selection unit B20 include, for example, a process signal indicating a combustion fuel flow rate, a process signal indicating a combustion air flow rate, a process signal indicating a fuel distribution ratio for diffusion combustion, a process signal indicating a fuel distribution ratio for premixed combustion, a process signal indicating an IGV angle which is the angle (opening) of the IGV, a process signal indicating a combustor bypass valve opening, a process signal indicating a gas turbine exhaust gas temperature, a process signal indicating a gas turbine blade path temperature, a process signal indicating an air compressor inlet temperature, a process signal indicating an air compressor outlet temperature, a process signal indicating an air compressor inlet pressure, and a process signal indicating combustion air humidity.

The process signal indicating the combustion fuel flow rate corresponds to, for example, the main fuel flow rate control valve position command value and the pilot fuel flow rate control valve position command value described in paragraph 0045 of Japanese Patent Application Laid-Open No. 2007-309279. The process signal indicating the combustion air flow rate corresponds to, for example, the input signal (input data) of the combustion air flow rate described in paragraph 0034 of Japanese Patent Application Laid-Open No. 2015-183619. The process signal indicating the fuel distribution ratio for diffusion combustion and the process signal indicating the fuel distribution ratio for premixed combustion correspond to, for example, the fuel distribution limit signal described in paragraph 0083 of Japanese Patent Application Laid-Open No. 10-127098.
The process signal indicating the IGV angle corresponds to, for example, the load command (load signal) described in paragraphs 0002 and 0003 of Japanese Patent Application Laid-Open No. 6-241062. The process signal indicating the combustor bypass valve opening corresponds to, for example, the combustor bypass valve opening command value described in paragraph 0045 of Japanese Patent Application Laid-Open No. 2007-309279.
The process signal indicating the gas turbine exhaust gas temperature corresponds to, for example, a signal measured by an exhaust gas thermometer described in paragraph 0023 of Japanese Patent Application Laid-Open No. 2016-142212. The process signal indicating the gas turbine blade path temperature corresponds to, for example, a signal measured by a blade path thermometer described in paragraph 0023 of Japanese Patent Application Laid-Open No. 2016-142212.

The process signal indicating the air compressor inlet temperature corresponds to, for example, the output signal of a temperature sensor measuring the compressor inlet temperature described in paragraph 23 of International Publication No. 2019/078309. The process signal indicating the air compressor outlet temperature corresponds to, for example, the output signal of a compressor outlet temperature sensor described in paragraph 17 of Japanese Patent Application Publication No. 2012-167550. The process signal indicating the air compressor inlet pressure corresponds to, for example, the output signal of an intake pressure meter described in paragraph 35 of Japanese Patent Application Publication No. 2016-142212. The process signal indicating the combustion air humidity corresponds to, for example, the output signal of a hygrometer installed on the suction side of the intake port of the intake duct described in paragraph 33 of Japanese Patent Application Publication No. 2020-002792, or the output signal of an inlet air humidity detector described in paragraph 21 of Japanese Patent Application Publication No. 2013-029095.

In the example shown in FIG. 6, the process signal selection unit B21 performs multiple regression analysis using the plurality of process signals P1, P2, P3, P4, P5, ..., Pi selected by the process signal selection unit B20 as explanatory variables and the concentration of nitrogen oxides at the inlet of the denitration device 7 as the response variable.
In addition, the process signal selection unit B21 selects a plurality of selected process signals P1, P2, P4, ..., Pi, which are process signals used to predict the concentration of nitrogen oxides at the inlet of the denitration device 7, from the plurality of process signals P1, P2, P3, P4, P5, ..., Pi selected by the process signal selection unit B20, based on the t value of the multiple regression analysis (for the t value, see, for example, the URL below) which indicates the degree of influence that the plurality of process signals P1, P2, P3, P4, P5, ..., Pi selected by the process signal selection unit B20 had on the concentration of nitrogen oxides at the inlet of the denitration device 7.
https://www.sangiin.go.jp/japanese/annai/chousa/keizai_prism/backnumber/r02pdf/202019202.pdf

The conformance evaluation unit B22 evaluates the conformance of the multiple selected process signals P1, P2, P4, ..., Pi selected by the process signal selection unit B21 based on the correction R2 of the multiple regression analysis (for details on the correction R2, see, for example, the URL below).
https://www.soumu.go.jp/ict_skill/pdf/ict_skill_3_4.pdf
In addition, the compatibility evaluation unit B22 determines multiple regression coefficients A, B, C, ..., i to be multiplied respectively to the multiple sorted process signals P1, P2, P4, ..., Pi sorted by the process signal sorting unit B21 in the multiple regression equation (nitrogen oxide concentration at the inlet of the denitration device 7 = P1 x A + P2 x B + P4 x C + ... + Pi x i + intercept).

In the example shown in FIG. 6, processing by the process signal selection unit B21 is performed once, and processing by the compatibility evaluation unit B22 is performed once, thereby selecting multiple selected process signals P1, P2, P4, ..., Pi and determining multiple regression coefficients A, B, C, ..., i. However, in other examples, processing by the process signal selection unit B21 and processing by the compatibility evaluation unit B22 may be repeated multiple times, thereby selecting multiple selected process signals P1, P2, P4, ..., Pi and determining multiple regression coefficients A, B, C, ..., i.

In the example shown in FIG. 6, the nitrogen oxide concentration prediction unit B23 predicts the concentration of nitrogen oxides at the inlet of the denitration device 7 based on the multiple sorted process signals P1, P2, P4, ..., Pi sorted by the process signal sorting unit B21, the multiple regression coefficients A, B, C, ..., i determined by the compatibility evaluation unit B22, and the intercept of the multiple regression equation.
Specifically, the nitrogen oxide concentration prediction unit B23 includes a multiplication unit B24, an intercept setting unit B25, and an addition unit B26. The multiplication unit B24 multiplies each of the plurality of sorted process signals P1, P2, P4, ..., Pi selected by the process signal selection unit B21 by each of the plurality of regression coefficients A, B, C, ..., i determined by the conformity evaluation unit B22. The intercept setting unit B25 sets an intercept of the multiple regression equation. The addition unit B26 adds the product (P1 x A, P2 x B, P4 x C, ..., Pi x i) obtained by the multiplication unit B24 to the intercept set by the intercept setting unit B25 to calculate the nitrogen oxide concentration at the inlet of the denitration device 7.

[Third embodiment]
A third embodiment of the denitration control device, the denitration control method, and the program according to the present invention will now be described.
The denitration control device 1 of the third embodiment is configured similarly to the denitration control device 1 of the first embodiment described above, except for the points described below. Therefore, the denitration control device 1 of the third embodiment can achieve the same effects as the denitration control device 1 of the first embodiment described above, except for the points described below.

The gas turbine plant GP to which the denitration control device 1 of the third embodiment is applied has the same configuration as the gas turbine plant GP to which the denitration control device 1 of the first embodiment shown in FIG. 1 is applied.
That is, in a gas turbine plant GP to which the denitration control device 1 of the third embodiment is applied, the denitration control device 1, an ammonia flow control valve 3, a chimney inlet nitrogen oxide concentration analyzer 4, a gas turbine 5, a flue 6, a denitration device 7, and a chimney 8 are installed, and the analyzer 2 (see FIG. 1 ) that measures the concentration of nitrogen oxides is not installed at the inlet of the denitration device 7.
The denitration control device 1 of the third embodiment controls the amount of ammonia supplied to the denitration device 7, similarly to the denitration control device 1 of the first embodiment.

In one example of a gas turbine plant GP to which the denitration control device 1 of the third embodiment is applied, similar to the example shown in FIG. 1 , the denitration control device 1 includes a nitrogen oxide concentration prediction model 11, a nitrogen oxide treatment amount calculation unit 12, an ammonia flow rate calculation unit 13, and a feedback correction amount calculation unit 14.
The nitrogen oxide concentration prediction model 11 of the denitration control device 1 of the third embodiment predicts the concentration of nitrogen oxides in the exhaust gas at the inlet of the denitration device 7 by model calculation, similar to the nitrogen oxide concentration prediction model 11 of the denitration control device 1 of the first embodiment.

FIG. 7 is a diagram showing an example of the nitrogen oxide concentration prediction model 11 of the denitration control device 1 of the third embodiment.
7, the nitrogen oxide concentration prediction model 11 is configured by a reaction rate model 11C. The reaction rate model 11C includes a fuel flow rate acquisition unit C30, a two-dimensional function unit C31, a combustion air amount acquisition unit C32, an IGV angle acquisition unit C33, a two-dimensional function unit C34, a divider C35, a multiplier C36, and a multiplier C37.

The fuel flow rate obtaining unit C30 obtains information indicating the fuel flow rate used for combustion in the combustor of the gas turbine 5. The information indicating the fuel flow rate obtained by the fuel flow rate obtaining unit C30 corresponds to, for example, a main fuel flow rate control valve position command value, a pilot fuel flow rate control valve position command value, or the like described in paragraph 0045 of Japanese Patent Application Laid-Open No. 2007-309279.
The two-dimensional function calculator C31 calculates a predicted value (FX [fuel flow rate]) of the reaction rate of the nitrogen oxide concentration in the exhaust gas at the inlet of the denitration device 7. In particular, the two-dimensional function calculator C31 calculates a predicted value (FX [fuel flow rate]) of the reaction rate of the nitrogen oxide concentration in the exhaust gas at the inlet of the denitration device 7 based on the fuel flow rate used for combustion in the combustor of the gas turbine 5 acquired by the fuel flow rate acquisition unit C30 and a predetermined relationship (predicted value = FX [fuel flow rate]) between the fuel flow rate and the predicted value of the reaction rate of the nitrogen oxide concentration in the exhaust gas at the inlet of the denitration device 7. The relationship (predicted value = FX [fuel flow rate]) between the fuel flow rate and the predicted value of the reaction rate of the nitrogen oxide concentration in the exhaust gas at the inlet of the denitration device 7 is determined in advance by, for example, conducting an experiment, a simulation, or the like.

The combustion air amount obtaining unit C32 obtains information indicating the combustion air amount, which is the amount of air used for combustion in the combustor of the gas turbine 5. The information indicating the combustion air amount obtained by the combustion air amount obtaining unit C32 corresponds to, for example, an input signal (input data) of the combustion air flow rate described in paragraph 0034 of JP 2015-183619 A.
The IGV angle acquisition unit C33 acquires information indicating the IGV angle. The information indicating the IGV angle acquired by the IGV angle acquisition unit C33 corresponds to, for example, the load command (load signal) described in paragraphs 0002 and 0003 of Japanese Patent Laid-Open No. 6-241062.
The two-dimensional function unit C34 calculates a reference air volume used to correct the predicted value of the reaction rate of the concentration of nitrogen oxides in the exhaust gas at the inlet of the denitration device 7, calculated by the two-dimensional function unit C31, based on the IGV angle acquired by the IGV angle acquisition unit C33. In detail, the two-dimensional function unit C34 calculates the reference air volume based on the IGV angle acquired by the IGV angle acquisition unit C33 and a preset relationship between the IGV angle and the reference air volume. The relationship between the IGV angle and the reference air volume is set in advance by, for example, conducting an experiment, a simulation, or the like.

The divider C35 calculates the air density used for combustion in the combustor of the gas turbine 5 by dividing the combustion air volume acquired by the combustion air volume acquisition unit C32 by the reference air volume calculated by the two-dimensional function unit C34.
The multiplier C36 multiplies the air density calculated by the divider C35 by a predetermined coefficient to calculate a value ([air density] ^α ) for correcting the predicted value (FX [fuel flow rate]) of the reaction rate of the nitrogen oxide concentration in the exhaust gas at the inlet of the denitration device 7 calculated by the two-dimensional function calculator C31.
Multiplier C37 multiplies the predicted value (FX [fuel flow rate]) of the reaction rate of the nitrogen oxide concentration in the exhaust gas at the inlet of the denitration device 7 calculated by the two-dimensional function calculator C31 by the value ([air density] ^α ) calculated by multiplier C36 to calculate the reaction rate of the nitrogen oxide concentration in the exhaust gas at the inlet of the denitration device 7 (d[NOx]/dt = [air density] ^α × FX [fuel flow rate]) and predicts the nitrogen oxide concentration in the exhaust gas at the inlet of the denitration device 7.
In other words, the reaction rate equation model 11C predicts the concentration of nitrogen oxides in the exhaust gas at the inlet of the denitration device 7 by correcting the predicted value of the reaction rate of the nitrogen oxide concentration (FX [fuel flow rate]) calculated by the two-dimensional function unit C31 based on the amount of combustion air used for combustion in the combustor of the gas turbine 5 and the IGV angle.

More specifically, in view of the fact that the generation of nitrogen oxides due to combustion in the combustor of the gas turbine 5 is a chemical reaction that depends on the air and combustion temperature, in the denitration control device 1 of the third embodiment, the nitrogen oxide concentration prediction model 11 predicts the concentration of nitrogen oxides in the exhaust gas at the inlet of the denitration device 7 by applying the reaction rate equations shown in the following equations (1) and (2) (Arrhenius equation).
d[NOx]/dt=k×[N ² ] ^α・[O ² ] ^β (1)
k=A×exp(-E/RT) (2)

The above formulas (1) and (2) can be expressed by the following formula (3).
d [NOx] / dt = A × combustion air concentration ^α × exp (-β / combustion temperature) (3)

When developing the reaction rate model 11C that predicts the concentration of nitrogen oxides in the exhaust gas at the inlet of the denitration device 7 from the above equation (3), the combustion air concentration is considered to be the air density, and the change in combustion temperature is considered to be a two-dimensional function based on the FX curve from the fuel flow rate (the above-mentioned FX [fuel flow rate]), and this can be expressed by the following equation (4).
d[NOx]/dt=[air density] ^α ×FX[fuel flow rate] (4)

The above-mentioned relationship used in the two-dimensional function calculator C31 to calculate the predicted value (FX [fuel flow rate]) of the reaction rate of nitrogen oxide concentration in exhaust gas at the inlet of the denitration device 7 is an FX curve obtained by plotting experimental results, simulation results, etc., with the fuel flow rate used for combustion in the combustor of the gas turbine 5 on the X axis and the predicted value of the concentration of nitrogen oxides produced by combustion in the combustor of the gas turbine 5 on the Y axis.

The above-mentioned relationship used to calculate the reference air volume in the two-dimensional function unit C34 was obtained by plotting experimental results, simulation results, etc. with the IGV angle on the X-axis and the reference air volume on the Y-axis.

[Fourth embodiment]
A fourth embodiment of the denitration control device, the denitration control method, and the program according to the present invention will now be described.
The denitration control device 1 of the fourth embodiment is configured similarly to the denitration control device 1 of the first embodiment described above, except for the points described below. Therefore, the denitration control device 1 of the fourth embodiment can achieve the same effects as the denitration control device 1 of the first embodiment described above, except for the points described below.

The gas turbine plant GP to which the denitration control device 1 of the fourth embodiment is applied has the same configuration as the gas turbine plant GP to which the denitration control device 1 of the first embodiment shown in FIG. 1 is applied.
That is, in a gas turbine plant GP to which the denitration control device 1 of the fourth embodiment is applied, the denitration control device 1, an ammonia flow control valve 3, a chimney inlet nitrogen oxide concentration analyzer 4, a gas turbine 5, a flue 6, a denitration device 7, and a chimney 8 are installed, and the analyzer 2 (see FIG. 1 ) that measures the concentration of nitrogen oxides is not installed at the inlet of the denitration device 7.
The denitration control device 1 of the fourth embodiment controls the amount of ammonia supplied to the denitration device 7, similarly to the denitration control device 1 of the first embodiment.

In one example of a gas turbine plant GP to which the denitration control device 1 of the fourth embodiment is applied, similar to the example shown in FIG. 1 , the denitration control device 1 includes a nitrogen oxide concentration prediction model 11, a nitrogen oxide treatment amount calculation unit 12, an ammonia flow rate calculation unit 13, and a feedback correction amount calculation unit 14.
The nitrogen oxide concentration prediction model 11 of the denitration control device 1 of the fourth embodiment predicts the concentration of nitrogen oxides in the exhaust gas at the inlet of the denitration device 7 by model calculation, similar to the nitrogen oxide concentration prediction model 11 of the denitration control device 1 of the first embodiment.

In the denitration control device 1 of the fourth embodiment, the nitrogen oxide concentration prediction model 11 includes at least two models selected from the group consisting of a three-dimensional function model 11A shown in FIG. 2, a multiple regression analysis model 11B shown in FIG. 6, and a reaction rate model 11C shown in FIG. 7.
In the denitration control device 1 of the fourth embodiment, the nitrogen oxide concentration prediction model 11 selects and uses one model from at least two models described above, or uses a combination of at least two models described above, depending on the characteristics of the gas turbine plant GP to which the denitration control device 1 of the fourth embodiment is applied.

Embodiments of the present invention have been described in detail above with reference to the drawings, but the specific configurations are not limited to these embodiments and examples, and modifications can be made as appropriate without departing from the spirit of the present invention. The configurations described in the above-mentioned embodiments and examples may also be combined.

Note that all or part of the functions of each unit of the denitration control device 1 in the above-described embodiment may be realized by recording a program for realizing these functions on a computer-readable recording medium, and reading and executing the program recorded on the recording medium into a computer system. Note that the term "computer system" here includes hardware such as an OS and peripheral devices.
Furthermore, "computer-readable recording media" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, as well as storage units such as hard disks built into computer systems. Furthermore, "computer-readable recording media" may also include devices that dynamically store programs for a short period of time, such as communication lines used when transmitting programs over networks like the Internet or communication lines like telephone lines, or devices that store programs for a fixed period of time, such as volatile memory within computer systems that serve as servers or clients in such cases. Furthermore, the above-mentioned programs may be programs that realize some of the aforementioned functions, or may be programs that can realize the aforementioned functions in combination with programs already stored in the computer system.

1...Denitrification control device, 11...Nitrogen oxide concentration prediction model, 11A...Three-dimensional function model, A10...Three-dimensional function unit, A11...Three-dimensional function unit, A12...Multiplier, A13...Subtractor, A14...Combustor bypass valve opening information acquisition unit, A15...Fuel/air ratio calculation unit, A16...Two-dimensional function unit, A17...Fixed value setting unit, A18...Switch, 11B...Multiple regression analysis model, B20...Process signal selection unit, B21...Process signal sorting unit, B22...Conformity evaluation unit, B23...Nitrogen oxide concentration prediction unit, B24...Multiplication unit, B25...Intercept setting B26...adder, 11C...reaction rate model, C30...fuel flow rate acquisition unit, C31...two-dimensional function calculator, C32...combustion air volume acquisition unit, C33...IGV angle acquisition unit, C34...two-dimensional function calculator, C35...divider, C36...multiplier, C37...multiplier, 12...nitrogen oxide treatment amount calculation unit, 13...ammonia flow rate calculation unit, 14...feedback correction amount calculation unit, 3...ammonia flow control valve, 4...stack inlet nitrogen oxide concentration analyzer, 5...gas turbine, 6...flue, 7...denitrification device, 8...stack, GP...gas turbine plant

Claims (8)

A denitration control device controls a supply amount of ammonia supplied to a denitration device that reduces the content of nitrogen oxides in exhaust gas containing nitrogen oxides generated by combustion in a gas turbine, the denitration control device comprising: a combustor in which premixed combustion and diffusion combustion are performed; and a combustor bypass valve that controls a fuel/air ratio, which is a ratio of fuel to air used for combustion in the combustor;
No analyzer for measuring the concentration of nitrogen oxides is installed at the inlet of the denitrification device,
a first three-dimensional function unit that calculates a nitrogen oxide concentration prediction value, the model being a first fuel amount that is a fuel amount used in the premixed combustion, a second fuel amount that is a fuel amount used in the diffusion combustion, and a predetermined first relationship between the first fuel amount, the second fuel amount, and a nitrogen oxide concentration prediction value that is a prediction value of the nitrogen oxide concentration;
a multiplier that corrects the predicted value of the nitrogen oxide concentration calculated by the first three-dimensional function unit based on a first nitrogen oxide concentration correction amount calculated according to the fuel/air ratio and an opening degree of the combustor bypass valve;
A denitrification control device comprising a nitrogen oxide concentration prediction model comprising : The nitrogen oxide concentration prediction model is
a second three-dimensional function unit that calculates the second nitrogen oxide concentration correction amount based on a combustion air temperature, which is the temperature of air used for combustion in the combustor, a combustion air humidity, which is the humidity of the air used for combustion in the combustor, and a second nitrogen oxide concentration correction amount that further corrects the nitrogen oxide concentration predicted value corrected by the multiplier,
The denitration control device according to claim 1 . The nitrogen oxide concentration prediction model is
a process signal selection unit that selects a plurality of process signals related to combustion in a combustor of the gas turbine, the process signals being assumed to affect the concentration of nitrogen oxides at the inlet of the denitration device;
a process signal selection unit that performs multiple regression analysis using the plurality of process signals selected by the process signal selection unit as explanatory variables and the concentration of nitrogen oxides as a response variable,
the process signal selection unit selects a plurality of selected process signals, which are a plurality of process signals used to predict the concentration of nitrogen oxides, from the plurality of process signals selected by the process signal selection unit based on a t value of a multiple regression analysis indicating the degree of influence of the plurality of process signals selected by the process signal selection unit on the concentration of nitrogen oxides;
The nitrogen oxide concentration prediction model is
a suitability evaluation unit that evaluates suitability of the plurality of selected process signals selected by the process signal selection unit based on a correction R2 of multiple regression analysis,
the compatibility evaluation unit determines a plurality of regression coefficients to be multiplied by the plurality of selected process signals selected by the process signal selection unit in a multiple regression equation;
The nitrogen oxide concentration prediction model is
a nitrogen oxide concentration prediction unit that predicts the concentration of nitrogen oxides at an inlet of the denitration device based on the plurality of sorted process signals selected by the process signal selection unit, the plurality of regression coefficients determined by the compatibility evaluation unit, and an intercept of the multiple regression equation;
The denitration control device according to claim 1 . In the combustor of the gas turbine, premixed combustion and diffusion combustion are performed,
the gas turbine is provided with a combustor bypass valve that controls a fuel/air ratio, which is a ratio of fuel to air used for combustion in the combustor;
The compressor of the gas turbine is provided with an inlet guide vane (IGV),
The plurality of process signals selected by the process signal selection unit include a plurality of process signals selected from a process signal indicating a combustion fuel flow rate, a process signal indicating a combustion air flow rate, a process signal indicating a fuel distribution ratio for diffusion combustion, a process signal indicating a fuel distribution ratio for premixed combustion, a process signal indicating an IGV angle, a process signal indicating a combustor bypass valve opening, a process signal indicating a gas turbine exhaust gas temperature, a process signal indicating a gas turbine blade path temperature, a process signal indicating an air compressor inlet temperature, a process signal indicating an air compressor outlet temperature, a process signal indicating an air compressor inlet pressure, and a process signal indicating combustion air humidity.
The denitration control device according to claim 3 . The compressor of the gas turbine is provided with an inlet guide vane (IGV),
The nitrogen oxide concentration prediction model is
a two-dimensional function calculator that calculates a predicted value of a reaction rate of the nitrogen oxide concentration based on a fuel flow rate used for combustion in a combustor of the gas turbine and a predetermined relationship between the fuel flow rate and a predicted value of a reaction rate of the nitrogen oxide concentration at an inlet of the denitration device,
The predicted value of the reaction rate of the nitrogen oxide concentration calculated by the two-dimensional function calculator is corrected based on the combustion air amount, which is the amount of air used for combustion in the combustor, and the IGV angle, which is the angle of the IGV, thereby predicting the nitrogen oxide concentration at the inlet of the denitration device.
The denitration control device according to claim 1 . In the combustor of the gas turbine, premixed combustion and diffusion combustion are performed,
the gas turbine is provided with a combustor bypass valve that controls a fuel/air ratio, which is a ratio of fuel to air used for combustion in the combustor;
The compressor of the gas turbine is provided with an inlet guide vane (IGV),
The nitrogen oxide concentration prediction model is
a first three-dimensional function unit that calculates the predicted nitrogen oxide concentration value based on a first fuel amount that is a fuel amount used in the premixed combustion, a second fuel amount that is a fuel amount used in the diffusion combustion, and a first predetermined relationship between the first fuel amount, the second fuel amount, and a predicted nitrogen oxide concentration value that is a predicted value of the nitrogen oxide concentration;
a multiplier that corrects the predicted value of the nitrogen oxide concentration calculated by the first three-dimensional function unit based on a first nitrogen oxide concentration correction amount calculated according to the fuel/air ratio and an opening degree of the combustor bypass valve;
a three-dimensional function model including a second three-dimensional function unit that calculates the second nitrogen oxide concentration correction amount based on a combustion air temperature that is the temperature of air used for combustion in the combustor, a combustion air humidity that is the humidity of the air used for combustion in the combustor, and a second nitrogen oxide concentration correction amount that further corrects the nitrogen oxide concentration predicted value corrected by the multiplier, and a second predetermined second relationship between the combustion air temperature, the combustion air humidity, and a second nitrogen oxide concentration correction amount;
a process signal selection unit that selects a plurality of process signals related to combustion in the combustor and that are assumed to affect the concentration of nitrogen oxides at the inlet of the denitration device;
a process signal selection unit that performs a multiple regression analysis using the plurality of process signals selected by the process signal selection unit as explanatory variables and the concentration of nitrogen oxides as a response variable, and selects a plurality of selected process signals from the plurality of process signals selected by the process signal selection unit, which are process signals used to predict the concentration of nitrogen oxides, based on a t value of the multiple regression analysis that indicates the degree of influence of the plurality of process signals selected by the process signal selection unit on the concentration of nitrogen oxides;
a compatibility evaluation unit that evaluates compatibility of the plurality of sorted process signals selected by the process signal selection unit based on a correction R2 of multiple regression analysis, and determines a plurality of regression coefficients to be multiplied by the plurality of sorted process signals selected by the process signal selection unit, respectively, in a multiple regression equation;
a multiple regression analysis model including a nitrogen oxide concentration prediction unit that predicts the concentration of nitrogen oxides at an inlet of the denitration device based on the plurality of selected process signals selected by the process signal selection unit, the plurality of regression coefficients determined by the compatibility evaluation unit, and an intercept of the multiple regression equation; and a two-dimensional function unit that calculates a predicted value of a reaction rate of the concentration of nitrogen oxides based on a fuel flow rate used for combustion in the combustor and a reaction rate equation that is a predetermined relationship between the fuel flow rate and a predicted value of a reaction rate of the concentration of nitrogen oxides at the inlet of the denitration device,
at least two reaction rate models that predict the concentration of nitrogen oxides at an inlet of the denitration device by correcting the predicted value of the reaction rate of the concentration of nitrogen oxides calculated by the two-dimensional function calculator based on a combustion air amount that is the amount of air used for combustion in the combustor and an IGV angle that is the angle of the IGV;
Selecting and using one of the at least two models, or using a combination of the at least two models, depending on the characteristics of the gas turbine plant to which the denitrification control device is applied.
The denitration control device according to claim 1 . A denitration control method for a denitration control device, comprising an ammonia supply amount control step of controlling a supply amount of ammonia supplied to a denitration device that reduces a content of nitrogen oxides in exhaust gas containing nitrogen oxides generated by combustion in a gas turbine that is equipped with a combustor in which premixed combustion and diffusion combustion are performed and a combustor bypass valve that controls a fuel/air ratio, which is a ratio of fuel to air used for combustion in the combustor,
No analyzer for measuring the concentration of nitrogen oxides is installed at the inlet of the denitrification device,
a nitrogen oxide concentration prediction step of predicting using a nitrogen oxide concentration prediction model that predicts the concentration of nitrogen oxides at the inlet of the denitration device, the nitrogen oxide concentration prediction model including: a first three-dimensional function unit that calculates the nitrogen oxide concentration prediction value based on a first fuel amount that is a fuel amount used in the premixed combustion, a second fuel amount that is a fuel amount used in the diffusion combustion, and a first predetermined relationship between the first fuel amount, the second fuel amount, and a nitrogen oxide concentration prediction value that is a prediction value of the nitrogen oxide concentration; and a multiplier that corrects the nitrogen oxide concentration prediction value calculated by the first three-dimensional function unit based on a first nitrogen oxide concentration correction amount calculated according to the fuel/air ratio and an opening degree of the combustor bypass valve.
Denitrification control method. The computer installed in the denitration control device
an ammonia supply amount control step of controlling a supply amount of ammonia supplied to a denitration device that reduces the content of nitrogen oxides in exhaust gases containing nitrogen oxides generated by combustion in a gas turbine that includes a combustor in which premixed combustion and diffusion combustion are performed and a combustor bypass valve that controls a fuel/air ratio, which is the ratio of fuel to air used for combustion in the combustor;
a first three-dimensional function unit that calculates the nitrogen oxide concentration predicted value based on a first fuel amount that is a fuel amount used in the premixed combustion, a second fuel amount that is a fuel amount used in the diffusion combustion, and a first predetermined relationship between the first fuel amount, the second fuel amount, and a nitrogen oxide concentration predicted value that is a predicted value of the nitrogen oxide concentration; and a multiplier that corrects the nitrogen oxide concentration predicted value calculated by the first three-dimensional function unit based on a first nitrogen oxide concentration correction amount calculated according to the fuel/air ratio and an aperture of the combustor bypass valve ,
No analyzer for measuring the concentration of nitrogen oxides is installed at the inlet of the denitrification device.
program.