Search

US-12623187-B2 - Exhaust gas treatment facility

US12623187B2US 12623187 B2US12623187 B2US 12623187B2US-12623187-B2

Abstract

A control system, for controlling an injection amount of a reducing agent injected into exhaust gas flowing from a coal-fired boiler in a thermal power generation facility toward a denitrification reactor of a denitrification device, includes: a first predictor predicting a first concentration of nitrogen oxides in the exhaust gas flowing toward the denitrification reactor based on first operation data of the thermal power generation facility; and a control device controlling the injection amount based on a predicted value of the first concentration. The first operation data includes at least either one of second operation data and third operation data, the second operation data being operation data of one or more coal pulverizers provided in the thermal power generation facility, and the third operation data being operation data of the coal-fired boiler affected by variation in operation conditions of the one or more coal pulverizers.

Inventors

  • Kazuo Yonekura
  • Takafumi Nishizu
  • Yasufumi Mochizuki
  • Hiroki Saitou

Assignees

  • IHI CORPORATION

Dates

Publication Date
20260512
Application Date
20200715
Priority Date
20190716

Claims (6)

  1. 1 . An exhaust gas treatment facility for treating exhaust gas discharged from a coal-fired boiler in a thermal power generation facility, the exhaust gas treatment facility comprising: a control valve controlling an injection amount of a reducing agent to be injected into exhaust gas flowing from the coal-fired boiler toward a denitrification reactor of a denitrification device; and a control system controlling a valve opening degree of the control valve, wherein the control system includes: a first predictor, based on a first operation data which is operation data of the thermal power generation facility at a first time point, predicting a first concentration, which is a concentration of nitrogen oxides in the exhaust gas flowing from the coal-fired boiler toward the denitrification reactor at a second time point after a predetermined period of time from the first operation data; a first calculator calculating a first needed flow rate which is a flow rate of the reducing agent at the second time point, needed to denitrify the nitrogen oxides such that a concentration of the nitrogen oxides is a target value or lower when the nitrogen oxides with the first concentration predicted by the first predictor are introduced into the denitrification device, and a controller controlling the valve opening degree of the control valve such that the injection amount at the second time point is based on a predicted value of the first concentration predicted by the first predictor and the first needed flow rate calculated by the first calculator, wherein the first operation data includes at least either one of a second operation data and a third operation data, the second operation data being operation data at the first time point of one or more coal pulverizers provided in the thermal power generation facility, and the third operation data being operation data at the first time point of the coal-fired boiler affected by variation in operation conditions of the one or more coal pulverizers.
  2. 2 . The exhaust gas treatment facility according to claim 1 , wherein the thermal power generation facility includes a plurality of the coal pulverizers, and the second operation data includes data indicating operation conditions of each of the plurality of the coal pulverizers.
  3. 3 . The exhaust gas treatment facility according to claim 2 , wherein the second operation data further includes supply amounts of coal supplied to the plurality of coal pulverizers and coal types of the coal supplied to the plurality of coal pulverizers.
  4. 4 . The exhaust gas treatment facility according to claim 1 , wherein the control system further includes: a first learner constructing a prediction model that outputs the predicted value of the first concentration, by performing machine learning using, as learning data, a data set of the first operation data and measured values of the first concentration obtained in past operations, wherein the first predictor inputs data including the first operation data at a predetermined time point to the prediction model constructed by the first learner and predicts the first concentration after a predetermined period of time has elapsed from the predetermined time point.
  5. 5 . The exhaust gas treatment facility according to claim 1 , wherein the control system further includes: a second predictor predicting a second concentration, which is a concentration of nitrogen oxides in the exhaust gas having passed through the denitrification reactor, based on a fourth operation data related to a denitrification reaction in the denitrification reactor; a subtractor calculating a difference value between a predicted value of the second concentration and the target value; a second calculator calculating a second needed flow rate which is a flow rate of the reducing agent needed for the denitrification device to denitrify all of the nitrogen oxides of the difference value calculated by the subtractor; and an adder calculating a third needed flow rate by adding the first needed flow rate and the second needed flow rate, and wherein the controller controls the valve opening degree of the control valve such that the injection amount is based on the third needed flow rate calculated by the adder, instead of the first needed flow rate.
  6. 6 . The exhaust gas treatment facility according to claim 5 , wherein the control system further includes: a second learner constructing a second prediction model that outputs a predicted value of the second concentration, by performing machine learning using, as learning data, a data set of the fourth operation data and measured values of the second concentration obtained in past operations, wherein the second predictor inputs data including the fourth operation data at a predetermined time point to the second prediction model constructed by the second learner and predicts the second concentration after a predetermined period of time has elapsed from the predetermined time point.

Description

The present application is a U.S. National Stage entry of International Application No. PCT/JP2020/027449, now WO 2021/010411, filed on Jul. 15, 2020. Priority is also claimed to Japanese Patent Application No. 2019-130924, filed on Jul. 16, 2019, the content of which is incorporated herein by reference. TECHNICAL FIELD The present disclosure relates to control systems. BACKGROUND Patent Document 1 below discloses a denitrification device that removes nitrogen oxides (NOx) in exhaust gas discharged from a boiler in a thermal power plant. The denitrification device introduces exhaust gas into a reactor including a catalyst and decomposes the nitrogen oxides in the exhaust gas into harmless nitrogen (N2) and water (H2O) by the action of the catalyst using a reducing agent (for example, ammonia (NH3)). DOCUMENT OF RELATED ART Patent Document [Patent Document 1] Japanese Unexamined Patent Application, First Publication No. 2014-211249 SUMMARY Technical Problem The flow rate of exhaust gas discharged from a boiler is not always constant and may vary greatly. Therefore, the operator of a thermal power plant sometimes sets the amount of the reducing agent to be injected into the exhaust gas (hereinafter referred to as “injection amount”) to a higher value such that the concentration of NOx emitted from the thermal power plant does not exceed the regulation value. Therefore, an excessive amount of reducing agent may be injected into the exhaust gas, and it is desirable to appropriately control the injection amount of the reducing agent even when the flow rate of the exhaust gas varies. The present disclosure has been made in view of such circumstances, and an object thereof is to provide a control system capable of appropriately controlling the injection amount of a reducing agent even when the flow rate of exhaust gas varies. Solution to Problem A control system of a first aspect of the present disclosure is a control system for controlling an injection amount of a reducing agent to be injected into exhaust gas flowing from a coal-fired boiler in a thermal power generation facility toward a denitrification reactor of a denitrification device, the control system including: a first predictor predicting a first concentration, which is a concentration of nitrogen oxides in the exhaust gas flowing toward the denitrification reactor, based on first operation data which is operation data of the thermal power generation facility; and a control device controlling the injection amount based on a predicted value of the first concentration predicted by the first predictor, wherein the first operation data includes at least either one of second operation data and third operation data, the second operation data being operation data of one or more coal pulverizers provided in the thermal power generation facility, and the third operation data being operation data of the coal-fired boiler affected by variation in operation conditions of the one or more coal pulverizers. A second aspect of the present disclosure is the control system of the first aspect, in which the thermal power generation facility includes a plurality of the coal pulverizers, and the second operation data includes data indicating operation conditions of each of the plurality of the coal pulverizers. A third aspect of the present disclosure is the control system of the second aspect, in which the second operation data further includes supply amounts of coal supplied to the plurality of coal pulverizers and coal types of the coal supplied to the plurality of coal pulverizers. A fourth aspect of the present disclosure is the control system of any one of the first to third aspects, in which the control system further includes: a first learner constructing a prediction model that outputs the predicted value of the first concentration, by performing machine learning using, as learning data, a data set of the first operation data and measured values of the first concentration obtained in past operations, wherein the first predictor inputs data including the first operation data at a predetermined time point to the prediction model constructed by the first learner and predicts the first concentration after a predetermined period of time has elapsed from the predetermined time point. A fifth aspect of the present disclosure is the control system of any one of the first to fourth aspects, in which the control device includes: a first calculator calculating a first needed flow rate which is a flow rate of the reducing agent needed to denitrify the nitrogen oxides such that a concentration of the nitrogen oxides is a target value or lower when the nitrogen oxides with the first concentration predicted by the first predictor are introduced into the denitrification device; and a controller controlling the injection amount based on the first needed flow rate calculated by the first calculator. A sixth aspect of the present disclosure is the control system of any one of the fi