Flue gas denitration system and control method thereof

Abstract

The invention discloses a flue gas denitration system and a control method thereof. The system comprises a control unit, a urea storage tank, a clear water storage tank, a urea delivery pump, a clear water delivery pump, an SNCR spray gun urea control valve A, an SNCR spray gun clear water control valve B, an ESNCR mechanical arm urea control valve C, an ESNCR mechanical arm clear water control valve D, an SNCR urea spray gun and an ESNCR denitration mechanical arm. The SNCR urea spray gun is located on the side wall of a hearth of the garbage incinerator. The ESNCR denitration mechanical arm is positioned in the middle of the hearth and can be lifted up and down; and the control unit is used for controlling the operation of the urea delivery pump and the clear water delivery pump, the opening degree of each control valve and the lifting and spraying angles of the ESNCR denitration mechanical arm. The method has the advantages of improving denitration efficiency, saving raw materials, reducing ammonia escape and the like.

Description

Technical Field

[0001] The present invention mainly relates to the technical field of waste incineration, and particularly relates to a flue gas denitration system and a control method thereof. Background Art

[0002] In the existing waste incinerator technology, SNCR (Selective Non-Catalytic Reduction) denitration is mostly used. Due to the limited denitration efficiency of SNCR, it is difficult to stably control nitrogen oxides below 100 mg / Nm³. Therefore, some waste incineration power plants adopt the strategy of combined SNCR and ESNCR denitration (ESNCR is a high-temperature resistant robotic arm denitration method. By extending the robotic arm into the middle of the furnace, monitoring the temperature change in the furnace, and finding the most suitable temperature point for efficient denitration) to control nitrogen oxides and ammonia slip not exceeding the standard limits.

[0003] In the combined use of SNCR and ESNCR systems in waste incinerators, how to allocate the denitration reducing agent has become a problem worthy of exploration. The traditional SNCR system sprays ammonia water or urea solution into the flue gas duct and reacts with NOx within a high-temperature window (usually 850 - 1100 °C). However, its control strategy is mostly fixed parameters or based on simple feedback, and has the following defects: 1. Static spraying strategy: It is unable to dynamically adjust the spraying amount and position according to real-time working conditions (such as flue gas flow rate, temperature fluctuation, NOx concentration change), resulting in low utilization rate of the reducing agent. Moreover, the denitration efficiency of simply using traditional SNCR spray guns is not high.

[0004] 2. Ammonia slip problem: Excessive spraying is likely to cause ammonia slip, increasing the operating cost and causing secondary pollution.

[0005] 3. Temperature sensitivity: When the temperature window shifts, it is difficult for traditional methods to timely adjust the spraying point, resulting in a sharp drop in denitration efficiency.

[0006] There are improved solutions (such as ESNCR, Efficient SNCR) to improve the efficiency through a denitration robotic arm (using a high-temperature resistant robotic arm to extend into the furnace, real-time sensing the flue gas temperature in the furnace area, making vertical position adjustments, finding the most suitable denitration reaction area, and adjusting the direction and spraying angle to make the urea solution fully mixed with the flue gas, and the urea spray covers most of the boiler cross-section to improve the denitration efficiency), but still lack the ability of dynamic global optimization. For example, in the process of combined use of SNCR and ESNCR, the allocation of the reducing agent is very important. If the reducing agent is unevenly distributed, it may lead to an increase in the local reducing agent concentration and cause ammonia slip to exceed the standard. At the same time, the dosage of the reducing agent is also very critical. Excessive dosage will not only cause waste of the reducing agent but also lead to an increase in ammonia slip. If the dosage is insufficient, it may lead to insufficient reaction of nitrogen oxides and increased emissions. Summary of the Invention

[0007] In view of the technical problems existing in the prior art, the present invention provides a flue gas denitration system and its control method that can improve denitration efficiency, save raw materials, and reduce ammonia escape.

[0008] To solve the above technical problems, the technical solution proposed by the present invention is as follows: A flue gas denitration system includes a control unit, a urea storage tank, a clean water storage tank, a urea transfer pump, a clean water transfer pump, a urea control valve A for the SNCR spray gun, a clean water control valve B for the SNCR spray gun, a urea control valve C for the ESNCR robotic arm, a clean water control valve D for the ESNCR robotic arm, an SNCR urea spray gun, and an ESNCR denitration robotic arm; the SNCR urea spray gun is located on the side wall of the furnace of the waste incinerator; the ESNCR denitration robotic arm is located in the middle of the furnace and can move up and down; The urea storage tank is connected to the SNCR urea spray gun successively through the urea transfer pump and the urea control valve A for the SNCR spray gun; The clean water storage tank is connected to the SNCR urea spray gun successively through the clean water transfer pump and the clean water control valve B for the SNCR spray gun; The urea storage tank is connected to the ESNCR denitration robotic arm successively through the urea transfer pump and the urea control valve C for the ESNCR robotic arm; The clean water storage tank is connected to the ESNCR denitration robotic arm successively through the clean water transfer pump and the clean water control valve D for the ESNCR robotic arm; The control unit is communicatively connected to the urea transfer pump, the clean water transfer pump, the urea control valve A for the SNCR spray gun, the clean water control valve B for the SNCR spray gun, the urea control valve C for the ESNCR robotic arm, the clean water control valve D for the ESNCR robotic arm, and the ESNCR denitration robotic arm respectively, and is used to control the operation of the urea transfer pump and the clean water transfer pump, the opening degree of each control valve, the lifting and spraying angle of the ESNCR denitration robotic arm; after adjusting the flow rate through the urea control valve A for the SNCR spray gun and the clean water control valve B for the SNCR spray gun, urea and water are mixed and then sent to the urea spray gun on the side wall of the furnace for denitration; through the urea control valve C for the ESNCR robotic arm and the clean water control valve D for the ESNCR robotic arm, urea and water are mixed and then sent to the central part of the furnace through the ESNCR denitration robotic arm for denitration.

[0009] Preferably, it further includes a data acquisition module, which acquires flue gas data, combustion parameters, and operating conditions data and sends them to the control unit. The control unit adjusts the opening degree of the urea control valve A for the SNCR spray gun, the clean water control valve B for the SNCR spray gun, the urea control valve C for the ESNCR robotic arm, and the clean water control valve D for the ESNCR robotic arm according to the flue gas data, combustion parameters, and operating conditions data, and adjusts the lifting height and spraying angle of the ESNCR denitration robotic arm to ensure the denitration efficiency.

[0010] Preferably, the source of flue gas data in the data acquisition module is the CEMS system of the waste incinerator.

[0011] Preferably, a dynamic prediction model is built in the control unit. The input of the dynamic prediction model is flue gas data, combustion parameters and operating conditions data, and the output is the optimal reductant distribution amount and fresh water consumption amount in the injection area. There is a mapping relationship between the flue gas data, combustion parameters, operating conditions data and the optimal reductant distribution amount and fresh water consumption amount in the injection area in the dynamic prediction model.

[0012] The present invention also discloses a control method based on the flue gas denitration system as described above, including the steps of: Obtain flue gas data, combustion parameters and operating conditions data and perform preprocessing to obtain preprocessed data; Input the preprocessed data as input data into a pre-built dynamic prediction model to obtain output results, including the optimal reductant distribution amount and fresh water consumption amount in each injection area; wherein there is a preset mapping relationship between the input data and the output results in the dynamic prediction model; Adjust the opening degrees of the urea control valve A of the SNCR spray gun, the fresh water control valve B of the SNCR spray gun, the urea control valve C of the ESNCR robotic arm, and the fresh water control valve D of the ESNCR robotic arm, and adjust the lifting and spraying angles of the ESNCR denitration robotic arm according to the optimal reductant distribution amount and fresh water consumption amount in each injection area.

[0013] Preferably, the dynamic prediction model is constructed based on a time series neural network of deep reinforcement learning.

[0014] Preferably, based on the dynamic prediction model, multi-objective optimization is carried out, and the denitration rate, ammonia slip rate and reductant consumption are balanced through a reinforcement learning algorithm; the reward function R is designed as: R = α * denitration rate - β * ammonia slip amount - γ * reductant consumption Wherein, α, β, and γ are the weight coefficients corresponding to the denitration rate, ammonia slip amount and reductant consumption respectively.

[0015] Preferably, the flue gas data includes NOx, O2, CO, temperature and flow rate; the combustion parameters include furnace temperature, primary air volume, secondary air volume and material layer thickness; the operating conditions data includes boiler load and waste calorific value.

[0016] Compared with the prior art, the advantages of the present invention are: The present invention realizes the dynamic allocation control of a reducing agent (such as urea) during the combined use of SNCR and ESNCR through a dynamic prediction model, enabling the combined use of SNCR and ESNCR to reach an optimal state, reducing the use of the reducing agent, while controlling nitrogen oxides not to exceed the emission limit and greatly reducing the ammonia slip value. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is an embodiment diagram of the flue gas denitration system of the present invention in specific applications.

[0018] Figure 2 It is a flowchart of the denitration control method of the present invention in an embodiment.

[0019] Legend: 1. Urea storage tank; 2. Fresh water storage tank; 3. Urea transfer pump; 4. Fresh water transfer pump; 5. SNCR urea spray gun; 6. ESNCR denitration robotic arm; 7. Furnace. DETAILED DESCRIPTION OF THE INVENTION

[0020] The present invention will be further described below in conjunction with the accompanying drawings of the specification and specific embodiments.

[0021] As Figure 1 shown, the flue gas denitration system provided by the embodiment of the present invention includes a control unit, a urea storage tank 1, a fresh water storage tank 2, a urea transfer pump 3, a fresh water transfer pump 4, an SNCR spray gun urea control valve A, an SNCR spray gun fresh water control valve B, an ESNCR robotic arm urea control valve C, an ESNCR robotic arm fresh water control valve D, an SNCR urea spray gun 5, and an ESNCR denitration robotic arm 6; the SNCR urea spray gun 5 is located on the side wall of the furnace 7 of the waste incinerator; the ESNCR denitration robotic arm 6 is located in the middle of the furnace 7 and can be lifted up and down; The urea storage tank 1 is connected to the SNCR urea spray gun 5 in sequence through the urea transfer pump 3 and the SNCR spray gun urea control valve A; The fresh water storage tank 2 is connected to the SNCR urea spray gun 5 in sequence through the fresh water transfer pump 4 and the SNCR spray gun fresh water control valve B; The urea storage tank 1 is connected to the ESNCR denitration robotic arm 6 in sequence through the urea transfer pump 3 and the ESNCR robotic arm urea control valve C; The fresh water storage tank 2 is connected to the ESNCR denitration robotic arm 6 in sequence through the fresh water transfer pump 4 and the ESNCR robotic arm fresh water control valve D; The control unit is communicatively connected to a urea delivery pump 3, a fresh water delivery pump 4, a urea control valve A of the SNCR spray gun, a fresh water control valve B of the SNCR spray gun, a urea control valve C of the ESNCR robotic arm, a fresh water control valve D of the ESNCR robotic arm, and an ESNCR denitration robotic arm 6 respectively, and is used to control the operation of the urea delivery pump 3 and the fresh water delivery pump 4, the opening degrees of each control valve, the lifting and spraying angle of the ESNCR denitration robotic arm 6, etc.; During the combined use of SNCR and ESNCR, traditional SNCR spray guns are used on the side walls of the boiler, and the ESNCR denitration robotic arm 6 is used for denitration at the central part of the furnace 7. The dosage of the reducing agent for the traditional SNCR spray gun and the dosage of the reducing agent for the ESNCR robotic arm are controlled by separately controlling the corresponding control valves. Specifically, after adjusting the flow rates through the urea control valve A of the SNCR spray gun and the fresh water control valve B of the SNCR spray gun, urea and water are mixed and then sent to the urea spray gun on the side wall of the furnace 7; wherein through the urea control valve C of the ESNCR robotic arm and the fresh water control valve D of the ESNCR robotic arm, urea and water are mixed and then sent to the central part of the furnace 7 through the ESNCR denitration robotic arm 6 for denitration.

[0022] Among them, ESNCR uses a high-temperature resistant robotic arm to achieve low-cost and high-efficiency denitration. This robotic arm can real-time sense the flue gas temperature in the furnace area, adjust the position in the vertical direction, and find the most suitable denitration reaction area; at the same time, this robotic arm has the ability to adjust the direction and spraying angle, so that the urea solution is fully mixed with the flue gas, and the urea spray covers most of the boiler cross-section, improving the denitration efficiency.

[0023] In a specific embodiment, it further includes a data acquisition module. The data acquisition module acquires flue gas data, combustion parameters, and operating conditions data and sends them to the control unit. The control unit adjusts the opening degrees of the urea control valve A of the SNCR spray gun, the fresh water control valve B of the SNCR spray gun, the urea control valve C of the ESNCR robotic arm, and the fresh water control valve D of the ESNCR robotic arm according to the flue gas data, combustion parameters, and operating conditions data, and adjusts the lifting height and spraying angle of the ESNCR denitration robotic arm 6 to ensure the denitration efficiency.

[0024] Specifically, a dynamic prediction model is built in the control unit. The input of the dynamic prediction model is flue gas data, combustion parameters, and operating conditions data, and the output is the optimal reducing agent distribution amount and fresh water usage amount in the spraying area. A mapping relationship between the flue gas data, combustion parameters, operating conditions data and the optimal reducing agent distribution amount and fresh water usage amount in the spraying area is set in the dynamic prediction model.

[0025] The present invention realizes the dynamic allocation control of a reducing agent (such as urea) during the combined use of SNCR and ESNCR through a dynamic prediction model, enabling the combined use of SNCR and ESNCR to reach an optimal state, reducing the use of the reducing agent, while controlling nitrogen oxides not to exceed the emission limit and greatly reducing the ammonia slip value.

[0026] As Figure 2 shown, an embodiment of the present invention further provides a flue gas denitrification control method, including the steps of: Collect data: Deploy on-line monitoring equipment (such as a CEMS system) to collect data such as NOx, O2, CO, temperature, and flow rate in the flue gas in real time; collect combustion parameters (furnace temperature, primary air volume, secondary air volume, bed layer thickness) and operating conditions data (boiler load, refuse calorific value); collect control parameters of the urea control valve A of the SNCR spray gun, the clean water control valve B of the SNCR spray gun, the urea control valve C of the ESNCR robotic arm, and the clean water control valve D of the ESNCR robotic arm; Realize the interconnection of each device through an industrial protocol (such as Modbus), construct a real-time database (such as a time series database) through each data, and then preprocess the data (denoising, normalization) and extract features; Construct a dynamic prediction model based on a time series neural network (such as LSTM) of deep reinforcement learning, input real-time data and historical data, and output the optimal reducing agent allocation amount and clean water usage amount of each injection area; then adjust the opening degrees of the control valves A-D according to the optimal reducing agent allocation amount and clean water usage amount of each injection area.

[0027] Specifically, the dynamic prediction model uses an LSTM network to predict the change trend of NOx concentration and the moving path of the temperature window in the next 5-10 minutes. When performing multi-objective optimization, the denitrification rate, ammonia slip rate, and reducing agent consumption are balanced through a reinforcement learning algorithm (such as PPO), and the reward function R is designed as: R = α * denitrification rate - β * ammonia slip amount - γ * reducing agent usage where α, β, and γ are the weight coefficients corresponding to the denitrification rate, ammonia slip amount, and reducing agent usage respectively, and the weight coefficients (α, β, γ) can be adjusted online adaptively.

[0028] In addition, it has an online learning mechanism: The system dynamically updates the model parameters according to the deviation between the actual denitrification effect and the predicted value to adapt to changes in fuel calorific value or equipment aging.

[0029] In addition, the AI calculation module focuses on the parameter calculation and training of the reducing agent control valves A, C and the clean water control valves B, D.

[0030] To better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings of the specification and specific implementation manners.

[0031] Taking the combined use and distribution of reducing agents in the SNCR / ESNCR intelligent system of a waste incineration power plant as an example: 1. Install an on-line continuous flue gas detector at the chimney to collect data such as NOx, O2, CO, temperature, and flow rate in the flue gas in real time, and collect the operation parameters of the DCS system of the waste incinerator in real time, such as the temperature of the furnace 7, the primary air volume, the secondary air volume, the thickness of the waste layer, the boiler load, the calorific value of the waste, the running speed of the grate, the opening of the urea control valve of the SNCR system, the opening of the clean water control valve of the SNCR system, the opening of the urea control valve of the mechanical arm of the ESNCR system, and the opening of the clean water control valve of the mechanical arm of the ESNCR system. The data sampling frequency is 1Hz; then transmit the data to the artificial intelligence analysis industrial control computer through Modbus communication; 2. Training stage: Collect historical data for 3 months (including data under different loads, calorific value changes of waste, different openings of urea control valves, and different openings of clean water control valves), and use the PyTorch framework to train the LSTM-DRL fusion model; 3. Control stage: The artificial intelligence analysis industrial control computer outputs data, and connects with the DCS system through Modbus communication to control the urea solution flow rate of 2 reducing agent control valves and the flow rate of 2 clean water control valves in real time; 4. Effect verification: Compared with the PID control of the traditional SNCR system, the denitrification efficiency is greatly improved, and the ammonia escape amount is significantly reduced.

[0032] In the denitrification process of the waste incinerator, the present invention realizes the intelligent distribution of reducing agents in the combined denitrification of the SNCR system and the ESNCR system, improves the denitrification efficiency, saves raw materials, and reduces ammonia escape at the same time.

[0033] The above is only the preferred implementation mode of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be pointed out that for those of ordinary skill in the art, several improvements and modifications made without departing from the principle of the present invention should be regarded as the protection scope of the present invention.

Claims

1. A flue gas denitration system, characterized in that, It includes a control unit, a urea storage tank (1), a clean water storage tank (2), a urea delivery pump (3), a clean water delivery pump (4), an SNCR spray gun urea control valve A, an SNCR spray gun clean water control valve B, an ESNCR robotic arm urea control valve C, an ESNCR robotic arm clean water control valve D, an SNCR urea spray gun (5), and an ESNCR denitration robotic arm (6); the SNCR urea spray gun (5) is located on the side wall of the furnace chamber (7) of the waste incinerator; the ESNCR denitration robotic arm (6) is located in the middle of the furnace chamber (7) and can move up and down; The urea storage tank (1) is connected to the SNCR urea spray gun (5) successively through the urea delivery pump (3) and the SNCR spray gun urea control valve A; The clean water storage tank (2) is connected to the SNCR urea spray gun (5) successively through the clean water delivery pump (4) and the SNCR spray gun clean water control valve B; The urea storage tank (1) is connected to the ESNCR denitration robotic arm (6) successively through the urea delivery pump (3) and the ESNCR robotic arm urea control valve C; The clean water storage tank (2) is connected to the ESNCR denitration robotic arm (6) successively through the clean water delivery pump (4) and the ESNCR robotic arm clean water control valve D; The control unit is communicatively connected to the urea delivery pump (3), the clean water delivery pump (4), the SNCR spray gun urea control valve A, the SNCR spray gun clean water control valve B, the ESNCR robotic arm urea control valve C, the ESNCR robotic arm clean water control valve D, and the ESNCR denitration robotic arm (6) respectively, and is used to control the operation of the urea delivery pump (3) and the clean water delivery pump (4), the opening degrees of each control valve, the lifting and spraying angle of the ESNCR denitration robotic arm (6); after adjusting the flow rates through the SNCR spray gun urea control valve A and the SNCR spray gun clean water control valve B, urea and water are mixed and then sent to the urea spray gun on the side wall of the furnace chamber (7) for denitration; through the ESNCR robotic arm urea control valve C and the ESNCR robotic arm clean water control valve D, urea and water are mixed and then sent to the center of the furnace chamber (7) through the ESNCR denitration robotic arm (6) for denitration.

2. The flue gas denitrification system according to claim 1, characterized in that, It further includes a data acquisition module, which acquires flue gas data, combustion parameters, and operating conditions data and sends them to the control unit. The control unit adjusts the opening degrees of the SNCR spray gun urea control valve A, the SNCR spray gun clean water control valve B, the ESNCR robotic arm urea control valve C, and the ESNCR robotic arm clean water control valve D according to the flue gas data, combustion parameters, and operating conditions data, and adjusts the lifting height and spraying angle of the ESNCR denitration robotic arm (6) to ensure the denitration efficiency.

3. The flue gas denitration system according to claim 2, wherein The source of the flue gas data in the data acquisition module is the CEMS system of the waste incinerator.

4. The flue gas denitration system according to claim 2 or 3, characterized in that, A dynamic prediction model is built in the control unit. The input of the dynamic prediction model is flue gas data, combustion parameters, and operating conditions data, and the output is the optimal reductant distribution amount and clean water consumption in the spraying area. A mapping relationship between the flue gas data, combustion parameters, operating conditions data and the optimal reductant distribution amount and clean water consumption in the spraying area is set in the dynamic prediction model.

5. A control method for a flue gas denitration system according to any one of claims 1-4, characterized in that, It includes steps: Obtain flue gas data, combustion parameters and operating conditions data and perform preprocessing to obtain preprocessed data; Use the preprocessed data as input data and input it into a pre-constructed dynamic prediction model to obtain output results, including the optimal reductant allocation amount and fresh water consumption of each injection zone; among them, there is a preset mapping relationship between the input data and the output results in the dynamic prediction model; Adjust the opening degrees of the urea control valve A of the SNCR spray gun, the fresh water control valve B of the SNCR spray gun, the urea control valve C of the ESNCR robotic arm, and the fresh water control valve D of the ESNCR robotic arm, and adjust the lifting and spraying angles of the ESNCR denitration robotic arm (6) according to the optimal reductant allocation amount and fresh water consumption of each injection zone.

6. The control method according to claim 5, wherein The dynamic prediction model is constructed based on a temporal neural network of deep reinforcement learning.

7. The control method according to claim 5 or 6, characterized in that, Based on the dynamic prediction model, perform multi-objective optimization, and balance the denitration rate, ammonia escape rate and reductant consumption through a reinforcement learning algorithm; the reward function R is designed as: R = α * denitration rate - β * ammonia escape amount - γ * reductant consumption where α, β, and γ are the weight coefficients corresponding to the denitration rate, ammonia escape amount, and reductant consumption, respectively.

8. The control method according to claim 5 or 6, characterized in that, The flue gas data includes NOx, O2, CO, temperature and flow rate; the combustion parameters include furnace temperature, primary air volume, secondary air volume and bed layer thickness; the operating conditions data include boiler load and refuse calorific value.

Images