A method for feed metering, calorific value identification, and combustion control in an iron powder boiler
By calculating the theoretical heat release of iron powder using a weightless metering and online detection device, and correcting the burnout coefficient using a recursive algorithm, the feeding and air distribution of the iron powder boiler were cascadedly regulated. This solved the problem of boiler load instability caused by the drift of iron powder calorific value, and improved the stability and safety of the control.
Patent Information
- Application Number
- CN202611451204.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-18
- Publication Date
- 2026-11-10
AI Technical Summary
Traditional feeder-based weighing or feed control based on fixed calorific value cannot identify the drift in calorific value of iron powder, leading to unstable boiler load control. This is especially true when there are fluctuations in metallization rate and particle size distribution, which can easily cause flameout or overheating.
A loss-in-weight metering device is used to collect the mass flow rate of iron powder in real time. Combined with the metallization rate and particle size distribution signals, the theoretical heat release per unit mass of iron powder is calculated by an online detection device. The burnout correction coefficient is corrected by a recursive algorithm to realize the cascade regulation of feeding and air distribution, and a control system with load feedforward and main steam pressure feedback is constructed.
It achieves real-time identification and stable control of the calorific value of iron powder, eliminates the disturbance of feed error to load control, ensures the safety and applicability of boiler under multi-burner and co-firing conditions, reduces load response time by 35%, and controls calorific value identification error within 5%.
Smart Images

Figure CN123064997A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of clean combustion and boiler automatic control technology, specifically relating to a method for feed metering and calorific value identification combustion control of iron powder boiler. Background Technology
[0002] Iron powder fuel has metallic iron as the main combustible component and iron oxide as the main combustion product. It has the characteristics of safe storage and transportation, easy recycling and reuse of combustion products, and strong coupling with existing coal-fired boilers. It is regarded as a clean solid fuel that can be co-fired or burned entirely.
[0003] Compared to coal, the calorific value per unit mass of iron powder mainly depends on the metallization rate, which fluctuates significantly due to the reduction process, oxidation during storage and transportation, and batch variations. Simultaneously, particle size distribution affects the degree of burnout, causing the actual calorific value of the iron powder entering the furnace to drift around the theoretical value of 7.0–7.5 MJ / kg (calculated as metallic iron). Traditional feeder-based weighing or feed control based on a fixed calorific value cannot identify this drift, leading to instability in main steam pressure, load response, and oxygen control, and in severe cases, causing flameout or overheating.
[0004] Existing boiler combustion control systems are mostly designed for pulverized coal, lacking online sensing of metallization rate and particle size distribution, as well as recursive identification and feed-air distribution coordination correction methods to address the uncertainty of iron powder calorific value. Therefore, an integrated method is needed that integrates metering at the feed end, calorific value identification at the fuel end, and cascade regulation at the control end. Summary of the Invention
[0005] The technical problem to be solved by this invention is to overcome the problem of fuel calorific value drift and boiler load control instability caused by fluctuations in the metallization rate of iron powder, and to provide a combustion control method that identifies the real-time calorific value of iron powder online based on feed metering and adjusts the feed and air distribution in a cascade manner accordingly.
[0006] To solve the above problems, the present invention adopts the following technical solution: a method for feed metering and calorific value identification combustion control of iron powder boiler, wherein the boiler controller executes steps S1 to S4. S1: Real-time acquisition of iron powder mass flow rate signals is achieved through a loss-in-weight metering device installed on the feed pipeline of each burner, and the metallization rate and particle size distribution signals of the iron powder entering the furnace are obtained through an online detection device. S2: Based on the reaction heat of iron oxidation to Fe2O3, the theoretical heat release per unit mass of iron powder is calculated according to the metallization rate, and a burnout correction coefficient is introduced in conjunction with the particle size distribution to obtain a soft measurement value of the fuel calorific value. S3: The boiler evaporation rate, furnace outlet oxygen volume fraction, and flue gas temperature are collected. Using the measured heat balance deviation as an observation, a recursive algorithm is used to continuously correct the burnout correction coefficient, and the real-time fuel calorific value is output. S4: Using the load command as feedforward and the main steam pressure and furnace outlet oxygen volume fraction as feedback, the target feed rate is calculated based on the real-time fuel calorific value, and cascade regulation is performed on the feed device, each secondary damper, and the burnout damper to ensure that the actual feed heat power follows the load command.
[0007] The theoretical heat release per unit mass of iron powder is calculated based on the enthalpy of the complete oxidation of metallic iron to Fe2O3, with a value of 7.0–7.5 MJ / kg, reduced according to the metallization rate. The metallization rate is measured online by an X-ray fluorescence analyzer or a magnetic susceptibility detector, and the particle size distribution is measured online by a laser particle size analyzer, with a detection cycle of 10–60 s. The burnout correction coefficient is obtained by regression of the product of D90 and specific surface area. The rolling correction adopts recursive least squares or Kalman filtering, with a correction cycle of 10–60 s and a relative identification error of no more than 5%. When multiple burners are connected in parallel, the heat distribution is balanced according to the real-time feed heat deviation, with the deviation in the same layer not exceeding ±5%. In the co-firing mode, the heat ratio is maintained according to the real-time calorific value and the low heating value of coal received. A minimum stable combustion load constraint and a conservative switching strategy for signal failure are also provided. Beneficial effects
[0008] (1) Introduce weightless metering, metallization rate and particle size online detection into the feeding process to directly obtain the true iron powder mass flow rate of each burner and eliminate the underlying disturbance of the load control caused by the feeding error.
[0009] (2) The burnout coefficient is recursively corrected by taking the enthalpy of the iron oxidation reaction as the physical basis and the thermal balance deviation as the observation, so that the relative error of real-time calorific value identification is no more than 5%, and the calorific value drift caused by the metallization rate fluctuation is suppressed.
[0010] (3) Load feedforward and main steam pressure and furnace outlet oxygen feedback constitute cascade regulation, and link each secondary air damper and burnout air damper to realize the rapid and stable follow of the feed heat power to the load command.
[0011] (4) It has multiple burner heat balance, iron and coal blending ratio closed loop, minimum stable combustion constraint and conservative handling of detection failure, which improves safety and applicability under full combustion and blending conditions. Attached Figure Description
[0012] Figure 1 This is a flowchart of the overall process of the present invention;
[0013] Figure 2 A schematic diagram of weight loss measurement, online detection, and burner arrangement;
[0014] Figure 3 A flowchart for soft measurement and recursive correction of fuel calorific value;
[0015] Figure 4 This is a block diagram of a load feedforward-pressure / oxygen feedback cascade control system.
[0016] Figure 5 This is a logic diagram for multi-burner balancing, blending ratio, and handling of abnormal operating conditions.
[0017] Explanation of reference numerals in the attached drawings: 1—Boiler body; 2—Furnace; 3—Burner; 4—Feeding pipeline; 5—Loss-in-weight metering device; 6—Feeding device; 7—Online metallization rate detection device; 8—Laser particle size analyzer; 9—Primary air; 10—Secondary air damper; 11—Burnout air damper; 12—Main steam pressure / evaporation rate measuring point; 13—Furnace outlet oxygen measuring point; 14—Flue gas temperature measuring point; 15—Controller; 16—Calorific value soft measurement module; 17—Recursive correction module; 18—Combustion control module; 19—Load command interface; 20—Coal feeding interface. Detailed Implementation
[0018] The present invention will be further described below with reference to the accompanying drawings. The embodiments described are for illustrative purposes only and are not intended to limit the scope of protection. Where there is no conflict, the technical features of the various embodiments can be combined with each other.
[0019] Example 1: Calorific value identification and load control in full-fire mode. For example... Figures 1 to 4 As shown, iron powder is transported to each burner (3) via the feed pipeline (4). The feed device (6) can be a screw feeder or a rotary feed valve, with a loss-in-weight metering device (5) connected in series upstream to output the mass flow rate F_i (t / h) in real time. The metallization rate online detection device (7) (X-ray fluorescence analyzer or magnetic susceptibility detector) and the laser particle size analyzer (8) are arranged at a representative section of the feed pipeline (4), with a detection cycle of 20 s.
[0020] The soft sensing module (16) for calorific value of the controller (15) calculates the theoretical heat release Q₀ = ΔH·m / 100 according to Equation (1), where ΔH is taken as 7.2 MJ / kg, and m is the metallization rate (%). The soft-sensed calorific value is calculated as Qc = Q₀·α according to Equation (2), and the burnout correction coefficient α = 1 - k·(D90-D₀)·SSA / SSA₀; during calibration, D90 = 45 μm and SSA₀ = 1.0 m² / g are taken as the benchmarks, k is taken as 0.004, an increase of 10 μm in D90 reduces α by about 0.02, and the calibration result is pre-stored in the controller.
[0021] The recursive correction module (17) constructs a heat balance deviation e = Psteam - Qc·ΣF_i using the boiler evaporation capacity D, the oxygen volume fraction O₂ at the furnace outlet, and the exhaust gas temperature Tgp, where Psteam is the heat absorption power converted from the evaporation capacity and the drum pressure. Recursive least squares with a forgetting factor λ = 0.98 is used to estimate α, or Kalman filtering is adopted with α as the state, and the observation equation is established based on the above heat balance; the correction period is 20 s, the single correction |Δα| is limited to ≤0.03, and the relative error of calorific value identification is controlled within 5%.
[0022] The combustion control module (18) uses the load command P* as feedforward, calculates the total feed according to the target feed rate F* = P* / Qc, the inner loop adjusts the rotation speed of the feeding device (6) through loss-in-weight metering feedback, and the outer loop corrects F* according to the main steam pressure deviation; the oxygen loop adjusts each secondary air damper (10) and burnout air damper (11) according to the set value of the oxygen volume fraction at the furnace outlet, with a control period of 1 to 5 s, forming a load-feed-air distribution cascade structure.
[0023] When multiple burners are arranged in parallel, balanced distribution is performed according to the deviation of the real-time feed heat Qc·F_i of each burner. When the deviation of the same layer is greater than ±5%, the feeding rotation speed is corrected burner by burner, and the corresponding secondary air dampers are finely adjusted synchronously to avoid local oxygen deficiency or slagging.
[0024] In the co-firing mode, coal is fed through the interface (20), and the controller maintains the heat ratio ρ = Qc·F_Fe / (Qnet,ar·F_coal + Qc·F_Fe) according to Equation (3), so that ρ is within a set range, for example, 30% to 70%; when the real-time calorific value of iron powder rises, F_Fe is reduced according to the calorific value equivalence principle, and F_Fe is increased otherwise.
[0025] In the minimum stable combustion constraint, the minimum stable combustion feed rate F_min is determined by cold state / hot state stable combustion tests; when F* < F_min, F is maintained at F_min, adjacent burners are put into operation or the stable combustion device is started, and further load reduction is blocked.
[0026] When the metallization rate or particle size signal fails, or the weight loss measurement deviation exceeds the limit, the controller will switch α to a conservative value α_safe (take the lower bias of the calibration envelope, for example, α_safe=0.90α0) and alarm; after the signal is restored, the recursive convergence will be restored, and the feed heat power deviation during the switching period will be compensated to the total feed command in an integral manner.
[0027] Example 2: Parameter Tuning. A 220 t / h pulverized iron boiler is equipped with 4 burners. The detection cycle is 20 s, the calorific value correction cycle is 20 s, the load control cycle is 2 s, and the oxygen control cycle is 2 s. ΔH is taken as 7.2 MJ / kg, k=0.004, λ=0.98, and the single correction limit is 0.03. In a disturbance test where the metallization rate drops from 88% to 82%, with relatively fixed calorific value control, the maximum deviation of the main steam pressure decreases from 0.55 MPa to 0.18 MPa, the fluctuation of the oxygen volume fraction at the furnace outlet decreases from ±2.1% to ±0.8%, and the load response time is shortened by approximately 35%.
[0028] Example 3: Abnormal operating conditions. The blending ratio is set to 50%±10%; when the laser particle size analyzer (8) malfunctions, switch to conservative α_safe and alarm, the iron coal feed is calculated according to the most recent effective calorific value, and the calorific value identification error is still no greater than 5% within 30 minutes of operation; when the target feed is lower than F_min, the system automatically retains the minimum stable combustion feed and prompts the operator to put the adjacent burner into operation to prevent flameout.
[0029] In the above embodiments, the enthalpy of Fe2O3 reaction, α regression formula, λ, limit and ratio range can all be modified through the controller parameter table after calibration; all equivalent substitutions, parameter tuning and simple modifications made within the spirit and principles of this invention fall within the protection scope of this invention.
Claims
1. A method for feed metering and calorific value identification combustion control in an iron powder boiler, characterized in that, The boiler controller performs the following steps: S1. The iron powder mass flow rate signal is collected in real time by a weightless metering device installed on the feed pipeline of each burner, and the metallization rate and particle size distribution signal of the iron powder entering the furnace are obtained by an online detection device. S2. Based on the reaction heat of the oxidation of metallic iron to produce Fe2O3, calculate the theoretical heat release per unit mass of iron powder according to the metallization rate obtained in step S1, and introduce a burnout correction coefficient in combination with particle size distribution to obtain the soft measurement value of fuel calorific value. S3. Collect boiler evaporation rate, furnace outlet oxygen volume fraction and flue gas temperature. Using the measured heat balance deviation as the observation, use a recursive algorithm to perform rolling correction on the burnout correction coefficient in step S2, and output the real-time fuel calorific value. S4. Using the power grid or heating network load command as the feedforward quantity and the main steam pressure and furnace outlet oxygen volume fraction as the feedback quantity, calculate the target feed rate according to the real-time fuel calorific value obtained in step S3, and perform cascade regulation on the feed device, each secondary damper and the burnout air damper so that the actual feed heat power follows the load command.
2. The method according to claim 1, characterized in that, The theoretical heat release per unit mass of iron powder mentioned in step S2 is calculated based on the enthalpy of the reaction of complete oxidation of metallic iron to Fe2O3, which is 7.0 to 7.5 MJ / kg. It is calculated based on metallic iron and reduced according to the metallization rate. After reduction, the theoretical heat release per unit mass of iron powder is Q0 = ΔH·m, where ΔH is the enthalpy of oxidation of metallic iron and m is the metallization rate.
3. The method according to claim 1, characterized in that, The metallization rate mentioned in step S1 is measured online by an X-ray fluorescence analyzer or a magnetic susceptibility detector installed on the feed pipeline, and the particle size distribution is measured online by a laser particle size analyzer with a detection cycle of 10 to 60 s; the iron powder particle size distribution is characterized by D90 and specific surface area SSA.
4. The method according to claim 1, characterized in that, The burnout correction coefficient α mentioned in step S2 is obtained by regression of the product of particle size distribution D90 and specific surface area SSA, α=1-k·(D90-D0)·SSA / SSA0, where k is determined by calibration experiment and pre-stored in the controller, and D0 and SSA0 are calibration reference values; and the reduction of burnout correction coefficient for every 10 μm increase of D90 is determined by calibration experiment and pre-stored in the controller.
5. The method according to claim 1, characterized in that, The rolling correction in step S3 adopts a recursive least squares algorithm or a Kalman filter algorithm, with a correction period of 10 to 60 s and a relative error of calorific value identification of no more than 5%. When the absolute value of the thermal balance deviation in multiple consecutive correction periods is less than the set threshold, the correction gain is reduced. When the deviation exceeds the set threshold, the correction gain is increased and the single correction amplitude is limited.
6. The method according to claim 1, characterized in that, In step S4, when multiple burners are running in parallel, the controller distributes the feed rate of each burner evenly according to the real-time feed heat deviation of each burner, so that the feed heat deviation of the burners in the same layer is no more than ±5%; the even distribution is based on the actual mass flow rate measured by the loss-in-weight metering device.
7. The method according to claim 1, characterized in that, In the iron powder and coal co-firing mode, the controller calculates the heat ratio of the two based on the real-time fuel calorific value and the received basis lower heating value of the coal, and maintains the co-firing heat ratio within the set range; when the real-time calorific value of the iron powder increases, the iron powder feeding rate is reduced according to the principle of calorific value equivalence, and vice versa.
8. The method according to claim 1, characterized in that, The controller is equipped with a minimum stable combustion load constraint. When the target feed rate calculated according to the load command is lower than the minimum stable combustion feed rate, the minimum stable combustion feed rate is maintained and the adjacent burner is put into operation or the stable combustion device is started, while the load reduction command is locked.
9. The method according to claim 1, characterized in that, When the metallization rate or particle size signal fails, the controller switches the burnout correction coefficient to a pre-calibrated conservative value and issues an alarm. The conservative value is taken as the side that makes the calculated calorific value too low. After the signal is restored, it re-converges according to the recursive algorithm and performs integral compensation on the feeding heat power during the switching period.
