Particle size-total iron collaborative monitoring method suitable for unit start-up stage of thermal power plant

Abstract

The application discloses a granularity-full iron collaborative monitoring method suitable for a unit start-up stage of a thermal power plant, which comprises the following steps: S1, basic data acquisition; S2, time series data preprocessing; S3, correlation relationship and trigger feature construction; S4, double-mode parallel monitoring determination, and two types of full iron detection control logics are synchronously operated; and S5, data closed-loop optimization. In the late start-up stage and the stable water quality stage, the gradient event trigger mechanism is relied on to automatically reduce the invalid detection times, the chemical reagent consumption is greatly reduced, the pump body and pipeline abrasion caused by frequent start and stop of online instruments is reduced, and the service life of the instruments is prolonged.

Description

Technical Field

[0001] This invention relates to the field of online monitoring technology for water vapor chemistry in thermal power plants, and more specifically, to a method for granularity and total iron co-monitoring based on an event-triggered gradient determination mechanism during the start-up operation of thermal power units. Background Technology

[0002] During the startup and heating process of thermal power units, large areas of oxide scale peel off from the inner walls of boiler pipes, and the concentration of iron oxide particles in the steam-water system fluctuates drastically. The total iron content in the steam-water system is a core indicator for judging whether the start-up water quality of the unit is up to standard. If the total iron content exceeds the standard for a long period of time, it can easily cause solid particle erosion of the steam turbine, scaling and under-deposit corrosion on the boiler heating surfaces, and seriously shorten the service life of the unit.

[0003] According to the current DL / T 561 "Guideline for Monitoring Water and Steam Chemistry in Thermal Power Plants," the monitoring of total iron in water and steam needs to be intensified during the unit start-up phase. Currently, the industry generally uses a fixed-cycle mode to control the detection frequency of online instruments for total iron, typically set to once every 2-3 hours. This monitoring method has significant shortcomings: The online iron analyzer uses intermittent sampling and measurement. Each test consumes chemical reagents and causes mechanical wear on the pump and pipeline. During periods of stable water quality, frequent and fixed tests result in significant waste of consumables and increase equipment maintenance costs. If the testing cycle is lengthened to save costs, a monitoring gap will appear, making it impossible to capture the sudden surge in iron content caused by the sudden peeling of oxide scale in time, which poses a safety hazard of missed detection in supervision; The water vapor particle counter can continuously output particle count data without interruption, and the total number of particles in water vapor has a significant positive correlation with the total iron content. However, currently, the particle counting monitoring and total iron detection systems operate independently, and the industry has not yet utilized continuous time-series particle data to dynamically adjust the frequency of total iron detection. Existing monitoring strategies rely solely on a single timing mechanism, lacking dynamic triggering logic that adapts to operating conditions. They fail to distinguish between two types of drastic water quality fluctuations: rapid rise and rapid fall of particles. This makes them unsuitable for the dual abrupt changes in water quality during the unit startup phase, which involve sudden increases and decreases in water quality, and makes it difficult to simultaneously meet the requirements of chemical monitoring safety and operational economy. Summary of the Invention

[0004] This invention provides a particle size-total iron co-monitoring method applicable to the start-up phase of thermal power plant units, in order to solve the technical problems existing in the prior art.

[0005] To achieve the above objectives, this invention provides a method for synergistic monitoring of particle size and total iron content suitable for the start-up phase of thermal power plant units, comprising: S1: Basic Data Collection Real-time continuous acquisition of particle counting time-series data output by water vapor particle counter and plotting of particle counting curve, establishment of communication link to connect to all-iron online analysis instrument, and input of unit start-up operating parameters and historical start-up water quality sample data of unit; S2, Time Series Data Preprocessing The particle counting time series data is processed by outlier removal, missing value completion and smoothing filtering to eliminate random noise caused by unit valve disturbance and water flow fluctuation; S3. Construction of Relationships and Triggering Features A mathematical model showing the positive correlation between the number of water vapor particles and the total iron content was fitted based on historical start-up water quality sample data of the unit; a sliding window average gradient model was constructed based on particle count time series data, the slope of the particle count curve was extracted as an event trigger feature, and the trigger thresholds for rapid particle rise and rapid particle fall were preset. S4. Dual-mode parallel monitoring and judgment, synchronously running two types of all-iron detection and control logic. The first monitoring mode is timed baseline measurement: a fixed duration forced cycle is set, which is not affected by the fluctuation of the particle counting curve. The online iron analyzer is periodically and automatically started to perform detection and to build a water vapor chemical monitoring safety baseline. The second monitoring mode is bidirectional gradient event-triggered measurement: continuously calculate the window average gradient of the particle counting curve in real time. When the window average gradient exceeds the corresponding preset threshold for M consecutive sampling times, an instruction is immediately issued to expedite the detection of total iron content. S5, Data Closed-Loop Optimization It summarizes all total iron detection data generated by timed benchmark measurements and bidirectional gradient event-triggered measurements, iteratively updates the positive correlation mathematical model between water vapor particle number and total iron content and the bidirectional gradient trigger threshold, and outputs complete water quality monitoring records and abnormal early warning information.

[0006] In one embodiment of the present invention, preferably, step S2, performing outlier removal, missing value completion, and smoothing filtering on the particle counting time series data, includes: identifying and removing outliers using the 3x standard deviation method, completing missing values ​​using linear interpolation, and performing smoothing filtering using a moving average algorithm.

[0007] In one embodiment of the present invention, preferably, in step S3, the mathematical model for the positive correlation between the number of water vapor particles and the total iron content adopts a linear regression model, and the parameters of the linear regression model are fitted based on multiple sets of historical water quality data from the start-up of the unit.

[0008] In one embodiment of the present invention, preferably, the second monitoring mode in step S4 is as follows: Single-point instantaneous gradient Calculation formula: , Where t is the current discrete sampling time, N(i) is the real-time total number of water vapor particles in the water vapor particle counter at time i, in units of particles / mL, and Δt is the fixed sampling time interval; A sliding window of length L is used to filter the gradient mean at a single point to obtain the final decision window average gradient. : , Where L is the moving average window length, Let be the average gradient of the window at time t, expressed in units of cells / (mL·min). This indicates a continuous decline in the number of particles.

[0009] In one embodiment of the present invention, preferably, the sampling interval Δt = 1 min, the sliding window length L = 3, and the number of consecutive sampling points M = 3.

[0010] In one embodiment of the present invention, preferably, in step S4, the determination condition for bidirectional gradient event-triggered measurement includes: Condition A: If the average gradient of the sampling window for M consecutive sampling times is greater than the threshold for triggering rapid particle rise, then the risk of large-scale oxide scale peeling off the pipeline is determined, and an expedited inspection is triggered. Condition B: If the average gradient of the sampling window for M consecutive sampling times is less than the threshold for triggering rapid particle descent, then the system is judged to be in a state of large-scale sewage discharge and rapid decline in water quality, and expedited detection is triggered. If either condition A or condition B is met, expedited testing will be initiated.

[0011] In one embodiment of the present invention, preferably, the monitoring instructions, water quality data, and exceedance warning signals output in step S5 are directly connected to the DCS chemical water vapor monitoring system of the thermal power plant to realize fully automatic closed-loop operation of the monitoring strategy.

[0012] In one embodiment of the present invention, preferably, in step S1, the particle counting time series data is the continuous online monitoring value of iron oxide particles in water vapor, which is used to predict the sudden change in total iron concentration caused by the peeling of pipe oxide scale during the start-up stage.

[0013] In one embodiment of the present invention, preferably, the particle size-all-iron collaborative monitoring method applicable to the start-up stage of thermal power plant units further includes a start-up phased adaptive threshold correction mechanism: the start-up process of the unit is divided into three sub-conditions: the temperature rise and ramp-up stage, the constant temperature stabilization stage, and the load reduction and sewage discharge stage, and independent particle rapid rise trigger thresholds and particle rapid fall trigger thresholds are stored for different sub-conditions. Based on the real-time unit heating rate and load parameters, the corresponding sub-condition particle rapid rise trigger threshold and particle rapid fall trigger threshold are automatically matched and dynamically replaced with the preset particle rapid rise trigger threshold and particle rapid fall trigger threshold in step S3.

[0014] In one embodiment of the present invention, preferably, the particle size-total iron co-monitoring method applicable to the start-up phase of thermal power plant units further includes a graded early warning and linkage control mechanism; the water quality early warning level is divided into three levels based on particle count time series data and the duration of the window average gradient exceeding the standard; the first-level early warning is only displayed in a pop-up window on the DCS interface; the second-level early warning simultaneously pushes an audible and visual alarm and shortens the subsequent timing reference measurement cycle; the third-level early warning links the unit's sewage valve and temperature regulation loop to output intervention commands, and simultaneously generates a start-up water quality exceeding the standard source tracing report, which fully records the particle gradient change, the total iron exceeding the standard time series, and the unit operating condition related data.

[0015] In one embodiment of the present invention, preferably, in step S4, the fixed duration forced period in the timing reference measurement is set to 4 hours.

[0016] The present invention has the following beneficial technical effects: Rigid safety boundary guarantee: 4-hour mandatory timed testing serves as a safety net, ensuring that there will be no long-term monitoring vacuum without testing data, regardless of whether the water quality is stable or not, and fully meeting the mandatory requirements of DL / T 561 water vapor chemical monitoring; Significantly reduce unit operation and maintenance costs: During the stable water quality phase in the mid-to-late stage of startup, the gradient event triggering mechanism automatically reduces the number of invalid detections, significantly reduces the consumption of chemical reagents, reduces the wear and tear on pumps and pipelines caused by frequent start-ups and shutdowns of online instruments, and extends the service life of instruments; Two-way fluctuation full-scenario sensitive response: Simultaneously identify two types of water quality change conditions: rapid rise and rapid fall of particles. Relying on the continuous data of the particle counter, it can detect severe fluctuation conditions such as oxide scale peeling and large-scale system sewage discharge in advance, and quickly start expedited detection to effectively avoid missed detection of conditions with excessive iron content. The noise reduction and false trigger prevention mechanism is improved: it adopts sliding window gradient filtering + continuous multi-point dual constraints to filter out single-point outliers of the instrument and random disturbances of water flow, which greatly reduces the frequency of false triggers. Adaptive iterative optimization capability: The gradient trigger threshold and particle-total iron correlation model can be continuously iterated and updated based on the unit's previous start-up data, adapting to start-up conditions of different capacities (300MW / 600MW / 1000MW) and different heating rates. It has strong on-site compatibility, requires no new hardware equipment, and can be deployed on the existing chemical online monitoring system simply by optimizing the DCS logic strategy. The workload of unit modification is small and the implementation cost is low. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of a particle size-total iron synergistic monitoring method applicable to the start-up phase of thermal power plant units, according to an embodiment of the present invention. Figure 2 This is a logical diagram of step S4 in an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Figure 1 This is a flowchart of a particle size-total iron synergistic monitoring method applicable to the start-up phase of thermal power plant units, according to an embodiment of the present invention. Figure 2 This is a logical diagram of step S4 according to an embodiment of the present invention, as follows: Figure 1 , Figure 2 As shown, the particle size-total iron synergistic monitoring method for thermal power plant unit start-up phase provided by the present invention includes: S1: Basic Data Collection Real-time continuous acquisition of particle counting time-series data output by water vapor particle counter and plotting of particle counting curve, establishment of communication link to connect to all-iron online analysis instrument, and input of unit start-up operating parameters and historical start-up water quality sample data of unit; In this embodiment, in step S1, the particle counting time series data is the continuous online monitoring value of iron oxide particles in water vapor, which is used to predict the sudden change in total iron concentration caused by the peeling of pipe oxide scale during the start-up stage.

[0021] S2, Time Series Data Preprocessing The particle counting time series data is processed by outlier removal, missing value completion and smoothing filtering to eliminate random noise caused by unit valve disturbance and water flow fluctuation; In this embodiment, step S2, which involves removing outliers, filling in missing values, and smoothing the particle counting time series data, includes: identifying and removing outliers using the 3x standard deviation method, filling in missing values ​​using linear interpolation, and using a moving average algorithm for smoothing filtering to ensure the stability and reliability of the time series curve.

[0022] S3. Construction of Relationships and Triggering Features A mathematical model showing the positive correlation between the number of water vapor particles and the total iron content was fitted based on historical start-up water quality sample data of the unit; a sliding window average gradient model was constructed based on particle count time series data, the slope of the particle count curve was extracted as an event trigger feature, and the trigger thresholds for rapid particle rise and rapid particle fall were preset. In this embodiment, in step S3, the positive correlation mathematical model between the number of water vapor particles and the total iron content adopts a linear regression model, and the parameters of the linear regression model are fitted based on multiple sets of historical water quality data from the start-up of the unit.

[0023] S4. Dual-mode parallel monitoring and judgment, synchronously running two types of all-iron detection and control logic. The first monitoring mode is timed baseline measurement: a fixed duration forced period (e.g., 4 hours) is set, which is not affected by the fluctuation of the particle counting curve. The online iron analyzer is periodically and automatically started to perform detection and to build a water vapor chemical monitoring safety baseline. The second monitoring mode is bidirectional gradient event-triggered measurement: continuously calculate the window average gradient of the particle counting curve in real time. When the window average gradient exceeds the corresponding preset threshold for M consecutive sampling times, an instruction is immediately issued to expedite the detection of total iron content. In this embodiment, the second monitoring mode in step S4 is as follows: Single-point instantaneous gradient Calculation formula: , Where t is the current discrete sampling time, N(i) is the real-time total number of water vapor particles in the water vapor particle counter at time i, in units of particles / mL, and Δt is the fixed sampling time interval; A sliding window of length L is used to filter the gradient mean at a single point to obtain the final decision window average gradient. : , Where L is the moving average window length, Let be the average gradient of the window at time t, expressed in units of cells / (mL·min). This indicates a continuous decline in the number of particles.

[0024] In this embodiment, the sampling interval Δt = 1 min, the sliding window length L = 3, and the number of consecutive sampling points M = 3.

[0025] In this embodiment, the determination criteria for bidirectional gradient event-triggered measurement include: Condition A: If the average gradient of the sampling window for M consecutive sampling times is greater than the threshold for triggering rapid particle rise, then the risk of large-scale oxide scale peeling off the pipeline is determined, and an expedited inspection is triggered. Condition B: If the average gradient of the sampling window for M consecutive sampling times is less than the threshold for triggering rapid particle descent, then the system is judged to be in a state of large-scale sewage discharge and rapid decline in water quality, and expedited detection is triggered. If either condition A or condition B is met, expedited testing will be initiated.

[0026] S5, Data Closed-Loop Optimization It summarizes all total iron detection data generated by timed benchmark measurements and bidirectional gradient event-triggered measurements, iteratively updates the positive correlation mathematical model between water vapor particle number and total iron content and the bidirectional gradient trigger threshold, and outputs complete water quality monitoring records and abnormal early warning information.

[0027] In this embodiment, the monitoring instructions, water quality data, and exceedance warning signals output in step S5 are directly connected to the DCS chemical water vapor monitoring system of the thermal power plant to realize the fully automatic closed-loop operation of the monitoring strategy.

[0028] In one embodiment of the present invention, preferably, the particle size-all-iron collaborative monitoring method applicable to the start-up stage of thermal power plant units further includes a start-up phased adaptive threshold correction mechanism: the start-up process of the unit is divided into three sub-conditions: the temperature rise and ramp-up stage, the constant temperature stabilization stage, and the load reduction and sewage discharge stage, and independent particle rapid rise trigger thresholds and particle rapid fall trigger thresholds are stored for different sub-conditions. Based on the real-time unit heating rate and load parameters, the corresponding sub-condition particle rapid rise trigger threshold and particle rapid fall trigger threshold are automatically matched and dynamically replaced with the preset particle rapid rise trigger threshold and particle rapid fall trigger threshold in step S3.

[0029] In one embodiment of the present invention, preferably, the particle size-total iron co-monitoring method applicable to the start-up phase of thermal power plant units further includes a graded early warning and linkage control mechanism; the water quality early warning level is divided into three levels based on particle count time series data and the duration of the window average gradient exceeding the standard; the first-level early warning is only displayed in a pop-up window on the DCS interface; the second-level early warning simultaneously pushes an audible and visual alarm and shortens the subsequent timing reference measurement cycle; the third-level early warning links the unit's sewage valve and temperature regulation loop to output intervention commands, and simultaneously generates a start-up water quality exceeding the standard source tracing report, which fully records the particle gradient change, the total iron exceeding the standard time series, and the unit operating condition related data.

[0030] This invention employs a dual-mode collaborative monitoring mechanism: a 4-hour timed baseline measurement and a particle timing slope event-triggered measurement. The timed measurement automatically initiates full iron detection every 4 hours to establish a monitoring baseline and safeguard the unit's chemical monitoring safety boundaries. It continuously calculates the slope of the particle count curve in real time, triggering expedited full iron detection when the slope exceeds a preset threshold. This invention reduces invalid detections and lowers operation and maintenance costs while accurately capturing instantaneous water quality anomalies during startup. The monitoring logic can be directly integrated into the power plant's existing DCS system, making it suitable for various types of thermal power units that frequently start and stop for peak shaving.

[0031] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of one embodiment, and the modules or processes shown in the drawings are not necessarily essential for implementing the present invention.

[0032] Those skilled in the art will understand that the modules in the apparatus of the embodiments can be distributed in the apparatus of the embodiments as described in the embodiments, or they can be located in one or more devices different from this embodiment with corresponding changes. The modules of the above embodiments can be combined into one module, or they can be further divided into multiple sub-modules.

[0033] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for synergistic monitoring of particle size and total iron content suitable for the start-up phase of thermal power plant units, characterized in that, include: S1: Basic Data Collection Real-time continuous acquisition of particle counting time-series data output by water vapor particle counter and plotting of particle counting curve, establishment of communication link to connect to all-iron online analysis instrument, and input of unit start-up operating parameters and historical start-up water quality sample data of unit; S2, Time Series Data Preprocessing The particle counting time series data is processed by outlier removal, missing value completion and smoothing filtering to eliminate random noise caused by unit valve disturbance and water flow fluctuation; S3. Construction of Relationships and Triggering Features A mathematical model showing the positive correlation between the number of water vapor particles and the total iron content was fitted based on historical start-up water quality sample data of the unit; a sliding window average gradient model was constructed based on particle count time series data, the slope of the particle count curve was extracted as an event trigger feature, and the trigger thresholds for rapid particle rise and rapid particle fall were preset. S4. Dual-mode parallel monitoring and judgment, synchronously running two types of all-iron detection and control logic. The first monitoring mode is timed baseline measurement: a fixed duration forced cycle is set, which is not affected by the fluctuation of the particle counting curve. The online iron analyzer is periodically and automatically started to perform detection and to build a water vapor chemical monitoring safety baseline. The second monitoring mode is bidirectional gradient event-triggered measurement: continuously calculate the window average gradient of the particle counting curve in real time. When the window average gradient exceeds the corresponding preset threshold for M consecutive sampling times, an instruction is immediately issued to expedite the detection of total iron content. S5, Data Closed-Loop Optimization It summarizes all total iron detection data generated by timed benchmark measurements and bidirectional gradient event-triggered measurements, iteratively updates the positive correlation mathematical model between water vapor particle number and total iron content and the bidirectional gradient trigger threshold, and outputs complete water quality monitoring records and abnormal early warning information.

2. The particle size-total iron co-monitoring method for the start-up phase of thermal power plant units according to claim 1, characterized in that, In step S2, the outlier removal, missing value completion, and smoothing filtering of the particle counting time series data include: using the 3x standard deviation method to identify and remove outliers, using linear interpolation to complete missing values, and using a moving average algorithm for smoothing filtering.

3. The particle size-total iron co-monitoring method for the start-up phase of thermal power plant units according to claim 1, characterized in that, In step S3, the mathematical model for the positive correlation between the number of water vapor particles and the total iron content adopts a linear regression model, and the parameters of the linear regression model are fitted based on multiple sets of historical water quality data from the start-up of the unit.

4. The particle size-total iron co-monitoring method for the start-up phase of thermal power plant units according to claim 1, characterized in that, The second monitoring mode in step S4 is as follows: Single-point instantaneous gradient Calculation formula: ， Where t is the current discrete sampling time, N(i) is the real-time total number of water vapor particles in the water vapor particle counter at time i, in units of particles / mL, and Δt is the fixed sampling time interval; A sliding window of length L is used to filter the gradient mean at a single point to obtain the final decision window average gradient. : ， Where L is the moving average window length, Let be the average gradient of the window at time t, expressed in units of cells / (mL·min). This indicates a continuous decline in the number of particles.

5. The particle size-total iron co-monitoring method for the start-up phase of thermal power plant units according to claim 4, characterized in that, The sampling interval Δt = 1 min, the sliding window length L = 3, and the number of consecutive sampling points M = 3.

6. The particle size-total iron synergistic monitoring method for the start-up phase of thermal power plant units according to claim 4, characterized in that, In step S4, the criteria for determining whether a bidirectional gradient event triggers a measurement include: Condition A: If the average gradient of the sampling window for M consecutive sampling times is greater than the threshold for triggering rapid particle rise, then the risk of large-scale oxide scale peeling off the pipeline is determined, and an expedited inspection is triggered. Condition B: If the average gradient of the sampling window for M consecutive sampling times is less than the threshold for triggering rapid particle descent, then the system is judged to be in a state of large-scale sewage discharge and rapid decline in water quality, and expedited detection is triggered. If either condition A or condition B is met, expedited testing will be initiated.

7. The particle size-total iron co-monitoring method for the start-up phase of thermal power plant units according to claim 1, characterized in that, The monitoring instructions, water quality data, and exceedance warning signals output in step S5 are directly connected to the DCS chemical water vapor monitoring system of the thermal power plant to realize the fully automatic closed-loop operation of the monitoring strategy.

8. The particle size-total iron co-monitoring method for the start-up phase of thermal power plant units according to claim 1, characterized in that, In step S1, the particle counting time series data is the continuous online monitoring value of iron oxide particles in water vapor, which is used to predict the sudden change in total iron concentration caused by the peeling of pipe oxide scale during the start-up stage.

9. The particle size-total iron co-monitoring method for the start-up phase of thermal power plant units according to claim 1, characterized in that, It also includes a phased adaptive threshold correction mechanism for startup: the startup process of the unit is divided into three sub-conditions: the temperature rise and ramp-up phase, the constant temperature and stabilization phase, and the load reduction and sewage discharge phase. Independent trigger thresholds for rapid particle rise and rapid particle fall are stored for different sub-conditions. Based on the real-time unit heating rate and load parameters, the corresponding sub-condition particle rapid rise trigger threshold and particle rapid fall trigger threshold are automatically matched and dynamically replaced with the preset particle rapid rise trigger threshold and particle rapid fall trigger threshold in step S3.

10. The particle size-total iron synergistic monitoring method for the start-up phase of thermal power plant units according to claim 1, characterized in that, It also includes a tiered early warning and linkage control mechanism; it divides water quality early warning levels into three levels based on particle count time series data and the duration of exceeding the average gradient in the window; Level 1 early warnings are only displayed in a pop-up window on the DCS interface; Level 2 early warnings simultaneously push out audible and visual alarms and shorten the subsequent timed reference measurement cycle; Level 3 early warnings link the unit's sewage valve and temperature regulation loop to output intervention commands, and simultaneously generate a start-up water quality exceeding the standard source report, which fully records particle gradient changes, total iron exceeding the standard time series, and unit operating condition related data.

Images