Graphite heat exchanger corrosion rate prediction system and method based on digital twinning

Abstract

The invention discloses a graphite heat exchanger corrosion rate prediction system and method based on digital twinning, and relates to the technical field of digital twinning, and the system comprises a corrosion rate prediction module which is used for carrying out linear mapping according to a comprehensive corrosion index and a preset reference corrosion rate, and calculating and outputting a real-time predicted corrosion rate of a graphite heat exchanger; and the maintenance decision support module is used for generating early warning information and a maintenance suggestion work order according to the corrosion risk level. According to the method, linear mapping calculation is carried out according to the comprehensive corrosion index, the preset reference corrosion rate and the reference comprehensive corrosion index through the corrosion rate prediction module, and the real-time predicted corrosion rate is output. The influence of medium, working condition and material state change on the corrosion rate can be dynamically reflected, and the real-time performance and accuracy of graphite heat exchanger corrosion rate prediction are greatly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of digital twinning, in particular to a graphite heat exchanger corrosion rate prediction system and method based on digital twinning. BACKGROUND

[0002] With the rapid development of process industries such as chemical industry, petrochemical industry, metallurgy, etc. towards large-scale, continuous, and high-parameter direction, heat exchange equipment, as the core device of energy transfer and medium temperature control in the production system, its running stability and service life are directly related to the efficiency, safety and economic benefits of the whole process. Graphite heat exchanger, with its excellent resistance to strong acid and strong base corrosion, good thermal conductivity and low thermal expansion coefficient, has shown irreplaceable advantages in handling high-corrosion working conditions such as chlorine-containing medium and concentrated acid, and has become a key core equipment in such industrial fields. Typical applications include chemical pickling, chlor-alkali production and wastewater deep treatment, etc. However, although graphite material has strong chemical stability, the dynamic changes of medium characteristics such as pH value fluctuation and chloride ion concentration change, working condition parameters such as medium temperature rise, flow rate anomaly and system pressure fluctuation, and material intrinsic state such as porosity increase and compressive strength decay with time will form a synergistic effect, continuously causing or accelerating corrosion behavior. Therefore, to realize real-time and accurate prediction of graphite heat exchanger corrosion rate and dynamic risk control has become one of the core needs to ensure the safe and stable operation of process industrial systems.

[0003] At present, the technical system for graphite heat exchanger corrosion monitoring and rate prediction is still dominated by traditional schemes, and the overall development is in the stage of offline detection combined with experience estimation. In the monitoring link, the existing technology relies on manual regular inspection and offline non-destructive testing after shutdown, such as ultrasonic testing and penetration testing. Although single type sensors such as only monitoring medium pH value or temperature are deployed in some scenarios, the data acquisition frequency is low, the monitoring dimension is limited, and it is difficult to capture the corrosion dynamic evolution process under the coupling action of multiple factors in real time. In the prediction link, the traditional method is mostly based on a single influencing factor to establish a linear empirical formula, such as only through medium pH value or temperature, or relying on historical failure data of the same type of equipment for analogy and inference, lacking systematic integration and deep coupling analysis of multi-source data such as medium corrosion, working condition erosion and material resistance, and unable to fully reflect the corrosion mechanism of the interaction of medium, working condition and material under complex working conditions. At the same time, although digital twinning technology has been gradually applied in the condition monitoring of high-end equipment such as aerospace and high-end machine tools, in the industrial equipment of graphite heat exchanger with specific material and complex corrosion environment, a complete technical chain of real-time acquisition of multi-source data, accurate prediction of coupled model, visualization of digital twinning and intelligentization of maintenance decision has not yet been formed, and the operation and maintenance personnel are still difficult to master the internal corrosion state of the equipment through intuitive digital carrier, and the maintenance decision lacks dynamic and accurate data support.

[0004] The prior art faces many outstanding problems in practical application, which seriously restricts the effectiveness of graphite heat exchanger corrosion control. First, the timeliness and continuity of monitoring are insufficient. Traditional offline detection needs to be implemented during shutdown, which not only interferes with the normal production rhythm, but also cannot cover the sudden corrosion risk in the shutdown gap, such as rapid pitting corrosion caused by sudden change of medium composition, which easily falls into the passive situation of no abnormality during detection and sudden failure during operation. Second, the prediction accuracy and reliability are low. Single factor model or fixed empirical formula cannot consider the synergistic effect of multi-dimensional parameters such as medium, working condition and material, especially in complex scenarios such as sudden increase of chloride ion concentration and abnormal increase of flow rate, the prediction result deviates significantly from the actual corrosion rate, which is difficult to support effective risk warning. Third, the visualization and decision support capability is weak. There is a lack of intuitive display means based on digital twinning, and the operation and maintenance personnel cannot quickly locate the high corrosion risk area of key parts such as tube sheet and heat exchange tube. Maintenance decisions are mostly based on manual experience, which is prone to problems such as excessive maintenance, such as replacing parts that have not reached the corrosion threshold, thereby increasing costs, or insufficient maintenance, such as delaying the repair of high-risk components, thereby leaving safety hazards. Fourth, the model adaptability and long-term effectiveness are poor. Existing prediction models mostly use fixed parameters, which cannot real-time correct weight factors and calculation parameters according to the dynamic changes of material performance during equipment aging process, such as the increase of porosity with service life and the attenuation of compressive strength. The prediction accuracy continues to decline after long-term operation, which is difficult to meet the corrosion dynamic management needs of graphite heat exchanger throughout its life cycle.

[0005] Therefore, it is necessary to invent a graphite heat exchanger corrosion rate prediction system and method based on digital twinning to solve the above problems. SUMMARY

[0006] The purpose of the present application is to provide a graphite heat exchanger corrosion rate prediction system and method based on digital twinning to solve the problems raised in the background art.

[0007] To achieve the above-mentioned purpose, the present application provides the following technical solution: a graphite heat exchanger corrosion rate prediction system based on digital twinning, comprising: A multi-source corrosion data acquisition module acquires medium characteristics, working condition parameters and body state data of the graphite heat exchanger through multi-sensors deployed on the graphite heat exchanger body and upstream and downstream process pipelines, and through non-destructive testing means during the maintenance period, forming a graphite heat exchanger multi-source corrosion data set; A corrosion influence coefficient analysis module is used to analyze the graphite heat exchanger multi-source corrosion data set to obtain medium corrosion coefficient, working condition corrosion coefficient and material resistance coefficient, and input the medium corrosion coefficient, working condition corrosion coefficient and material resistance coefficient into the graphite corrosion coupling prediction model, and output the comprehensive corrosion index; The corrosion rate prediction module is configured to calculate and output a real-time predicted corrosion rate of the graphite heat exchanger by linear mapping of the comprehensive corrosion index in combination with a preset reference corrosion rate; The digital twin visualization module is configured to receive the real-time predicted corrosion rate, the comprehensive corrosion index, and the graphite heat exchanger multi-source corrosion dataset, drive a three-dimensional digital twin of the graphite heat exchanger to perform visual display, and mark a corrosion risk level. The maintenance decision support module is configured to generate early warning information and a maintenance suggestion work order according to the corrosion risk level.

[0008] Preferably, the graphite heat exchanger multi-source corrosion dataset includes medium characteristic data, operating condition data, and body state data; the medium characteristic data includes a medium pH value C PH , a chloride ion concentration C; the operating condition data includes a medium temperature T, a medium flow rate V, and a system pressure P; and the body state data includes a graphite material porosity P r , a graphite material compressive strength σ c .

[0009] Preferably, the medium corrosivity coefficient is specifically: , wherein C PH is the medium pH value, C is the chloride ion concentration, θ PH is a pH corrosion critical threshold, γ is a chloride ion concentration saturation constant, k PH is a pH sensitivity coefficient, α and β are weight factors and satisfy α+β=1, and e is a natural constant.

[0010] Preferably, the operating condition corrosivity coefficient is specifically: , wherein T is the medium temperature, T0 is a reference reference value of the medium temperature, V is the medium flow rate, V0 is a reference reference value of the medium flow rate, P is the system pressure, P0 is a reference reference value of the system pressure, and η is a sensitivity adjustment index.

[0011] Preferably, the material resistance coefficient is specifically: , wherein P r is a measured graphite porosity, P r0 is a standard porosity of a new material, σ c is a measured graphite compressive strength, σ c0 is a standard compressive strength of a new material, and λ is a porosity influence amplification coefficient.

[0012] Preferably, the graphite corrosion coupling prediction model is specifically: , wherein Q is a comprehensive corrosion index, F is a medium corrosivity coefficient, E is a working condition corrosiveness coefficient, M is a material resistance coefficient, ω1, ω2 and ω3 are dynamic weight factors and satisfy ω1+ω2+ω3=1.

[0013] Preferably, the calculation formula of the real-time predicted corrosion rate is: , wherein V predicted is the real-time predicted corrosion rate, V base is a reference corrosion rate, Q is a comprehensive corrosion index, and Q base is a reference comprehensive corrosion index.

[0014] Preferably, the corrosion risk level is: When the real-time predicted corrosion rate V predicted >V1, it is a first-class corrosion risk; When the real-time predicted corrosion rate V2<V predicted ≤V1, it is a second-class corrosion risk; When the real-time predicted corrosion rate V3<V predicted ≤V2, it is a third-class corrosion risk; When the real-time predicted corrosion rate V predicted ≤V3, it is a fourth-class corrosion risk.

[0015] Preferably, the maintenance decision support module is further configured to perform model dynamic correction, comprising: a bias quantification unit for calculating a relative error coefficient δ according to an actual corrosion rate V actual obtained by non-destructive testing and a real-time predicted corrosion rate V predicted ; ; a correction triggering unit for starting correction when the relative error coefficient δ exceeds a preset threshold δ th and the actual corrosion rate V actual exceeds a minimum effective corrosion rate threshold V min ; a weight adjustment unit for updating the weight factor by using an exponential decay method: , wherein, is a current weight factor, is an updated weight factor, μ is a learning rate, δ is a relative error coefficient, e is a natural constant, is a weight adjustment direction function, wherein, wherein, To predict the corrosion rate, the partial derivative of the weight ω i ; A constraint processing unit is configured to normalize the dynamic weight factors ω1, ω2 and ω3 and apply boundary constraints; A freeze control unit is configured to suspend the correction when the average relative error of the last 5 corrections is lower than a preset error threshold.

[0016] The graphite heat exchanger corrosion rate prediction method based on digital twinning specifically comprises the following steps: S1, a multi-source corrosion data acquisition module acquires the medium characteristics, operating parameter and body state data of the graphite heat exchanger through a plurality of sensors deployed on the graphite heat exchanger body and upstream and downstream process pipelines and through non-destructive testing means during the maintenance period to form a multi-source corrosion data set of the graphite heat exchanger; S2, a corrosion influence coefficient analysis module is configured to analyze the multi-source corrosion data set of the graphite heat exchanger to obtain a medium corrosivity coefficient, an operating corrosivity coefficient and a material resistance coefficient, and input the medium corrosivity coefficient, the operating corrosivity coefficient and the material resistance coefficient into a graphite corrosion coupling prediction model to output a comprehensive corrosion index; S3, a corrosion rate prediction module is configured to perform linear mapping according to the comprehensive corrosion index in combination with a preset reference corrosion rate, calculate and output a real-time predicted corrosion rate of the graphite heat exchanger; S4, a digital twinning visualization module is configured to receive the real-time predicted corrosion rate, the comprehensive corrosion index and the multi-source corrosion data set of the graphite heat exchanger, drive a three-dimensional digital twin of the graphite heat exchanger to perform visual display and mark a corrosion risk level; S5, a maintenance decision support module is configured to generate early warning information and maintenance suggestion work orders according to the corrosion risk level.

[0017] Technical effects and advantages of the present application: 1, the present application analyzes the multi-source corrosion data set, respectively calculates the medium corrosivity coefficient, the operating corrosivity coefficient and the material resistance coefficient, and inputs the three into the graphite corrosion coupling prediction model with dynamic weight factors to output the comprehensive corrosion index, realizes accurate quantification and multi-factor coupling analysis of the corrosion influencing factors, breaks through the limitations of traditional single factor evaluation, and can more truly reflect the actual corrosion mechanism of the graphite heat exchanger; 2, the present application uses the corrosion rate prediction module to perform linear mapping calculation according to the comprehensive corrosion index, the preset reference corrosion rate and the reference comprehensive corrosion index, and outputs the real-time predicted corrosion rate, compared with the traditional experience estimation method, this method based on multi-source data and coupling model can dynamically reflect the influence of the medium, operating condition and material state change on the corrosion rate, greatly improves the real-time performance and accuracy of the graphite heat exchanger corrosion rate prediction; 3. This invention achieves a closed-loop iteration of "prediction-deviation analysis-model optimization" by maintaining the decision support module configuration model dynamic correction function. It can continuously optimize the dynamic weights of the coupled prediction model according to the actual corrosion rate, avoid prediction deviations caused by changes in operating conditions or equipment aging, and maintain high prediction accuracy in the long term. 4. This invention divides the real-time predicted corrosion rate into four levels of corrosion risk and matches different early warning priorities and maintenance suggestion work orders to different risk levels, thereby achieving refined and differentiated management and control of corrosion risk. This avoids resource waste or high-risk omissions caused by a uniform operation and maintenance strategy, and improves the scientific nature of operation and maintenance decisions and the efficiency of resource utilization. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the system structure of the present invention.

[0019] Figure 2 This is a schematic diagram of the method steps of the present invention. Detailed Implementation

[0020] 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.

[0021] This invention provides, for example Figure 1 The digital twin-based graphite heat exchanger corrosion rate prediction system shown includes: The multi-source corrosion data acquisition module collects medium characteristics, operating parameters and body status data of the graphite heat exchanger by deploying multiple sensors on the graphite heat exchanger body and upstream and downstream process pipelines, as well as by non-destructive testing during the maintenance period, and forms a multi-source corrosion dataset of the graphite heat exchanger. Furthermore, in the above technical solution, the multi-source corrosion dataset of the graphite heat exchanger includes media characteristic data, operating condition data, and body state data; the media characteristic data includes the media pH value C. PH The chloride ion concentration C; the operating data includes medium temperature T, medium flow rate V, and system pressure P; the bulk condition data includes graphite material porosity P. r σ, compressive strength of graphite materials c .

[0022] It is important to know that the pH value C PHThe corrosion-resistant online PH sensor is deployed for collection at the medium inlet and outlet pipe sections of the graphite heat exchanger body, the key points inside the heat exchange chamber, such as the connection between the tube sheet and the heat exchange tube, and the medium conveying sections of the upstream and downstream process pipelines, such as the feed pipe and the reflux pipe. The selected sensor needs to adapt to the corrosion, temperature and pressure conditions of the system medium. For example, a glass electrode type PH sensor with a polytetrafluoroethylene shell is selected. The sensor probe directly contacts the medium, collects PH value data in real time, and sends them to the data acquisition and twin construction module through wired or wireless transmission. The collection frequency is set to 1 time / minute to capture the dynamic changes of the medium PH value. The sensor needs to be calibrated on site once a month using a standard PH buffer solution to adjust the zero point and slope, ensuring that the measurement error is controlled within ±0.05 PH units; The chloride ion concentration C is collected by deploying an online chloride ion concentration sensor at the tube side and shell side medium outlet pipelines of the graphite heat exchanger body, and the medium sampling sections of the upstream and downstream process pipelines. The sensor type is selected to match the ion-selective electrode sensor with the characteristics of the medium, and a temperature compensation module is equipped to offset the influence of medium temperature changes on the measurement results. When installing the sensor, the probe needs to be completely immersed in the medium and away from the bubble generation area. Real-time chloride ion concentration data are collected and transmitted to the data acquisition and twin construction module. The collection frequency is 1 time / 5 minutes. A standard chloride ion solution with known concentration is used to calibrate the sensor every quarter to ensure that the measurement accuracy meets the requirement of ±1 mg / L. At the same time, the probe surface is regularly cleaned of dirt and deposits to avoid affecting the ion response speed. The medium temperature T is collected by deploying a platinum resistance temperature sensor or a thermocouple temperature sensor at the outer wall of the heat exchange tube in the middle of the tube side inlet and outlet and shell side inlet and outlet pipelines of the graphite heat exchanger body, and the straight pipe section before the valve after the pump of the upstream and downstream process pipelines. The sensor needs to be packaged with a corrosion-resistant metal protective sleeve. When installing, the sleeve is inserted into the medium flow channel with an insertion depth not less than 1 / 3 of the pipe diameter to ensure full contact with the medium. Real-time temperature data are collected and transmitted to the data acquisition and twin construction module. The collection frequency is 1 time / 30 seconds to track temperature fluctuations in real time. Before installation, the sensor needs to be temperature calibrated through constant temperature oil bath or dry furnace in the-20℃~200℃ interval. The calibration status is reviewed once every half year during operation to ensure that the measurement error does not exceed ±0.5℃; The medium flow rate V is collected by deploying an electromagnetic flowmeter or an ultrasonic flowmeter on the straight pipe section of the upstream and downstream process pipelines of the inlet and outlet pipes of the tube side and shell side of the graphite heat exchanger body. The flowmeter type is determined according to the corrosivity and viscosity of the medium and the pipe diameter. When the electromagnetic flowmeter is installed, the full pipe flow of the pipeline needs to be ensured, the sensor electrode axis is perpendicular to the direction of the medium flow and is located at the center axis of the pipeline. The ultrasonic flowmeter is installed in an external clamping mode to avoid damaging the integrity of the pipeline. The probes are installed on the horizontal diameter of the outer wall of the pipeline on both sides to ensure that the ultrasonic signal can effectively penetrate the pipe wall and propagate in the medium. The flow rate data is collected in real time and transmitted to the data acquisition and twin construction module. The collection frequency is 1 time / 1 minute. The flowmeter is calibrated every quarter using a standard flow device to ensure that the flow rate measurement error is controlled within ±1%. The system pressure P is collected by deploying a pressure sensor on the pressure key points such as the inlet of the pump outlet valve and the outlet of the storage tank of the upstream and downstream process pipelines of the pressure interface of the tube side and shell side of the graphite heat exchanger body. The sensor type is selected as a piezoelectric or strain gauge pressure sensor. The range needs to cover 1.2~1.5 times the normal working pressure of the system and has the functions of resisting medium corrosion and temperature compensation. The sensor is installed at the pressure interface through threaded or flanged connection to ensure good sealing and no leakage. The pressure data is collected in real time and transmitted to the data acquisition and twin construction module. The collection frequency is 1 time / 1 minute. Before installation, the 0 point and full-scale calibration need to be performed through a standard pressure source. The zero drift is checked once a month during operation to ensure that the pressure measurement error is not more than ±0.2% FS. The porosity P of the graphite material r During the maintenance shutdown of the graphite heat exchanger, the porosity P of the graphite material is collected by using non-destructive testing methods combined with sampling detection. Specifically, for the key structural parts such as the tube plate shell of the heat exchange tube of the body, the ultrasonic detection method is preferred. A 2.5MHz~5MHz straight probe is used to apply coupling agent on the surface of the graphite to avoid corrosion of the graphite. The ultrasonic wave is emitted and the internal reflected signal is received. According to the change of ultrasonic propagation speed and reflected amplitude, combined with the pre-established porosity-ultrasonic velocity calibration curve, the measured porosity P is calculated. r ; The compressive strength σ of the graphite material c According to the national standard GB / T 13465.2-2014, a universal material testing machine is used to perform compression test on a sample processed into a standard size such as 20mm×20mm×20mm. The testing machine applies pressure at a constant loading rate of 1.5mm / min until the sample is destroyed. The maximum pressure value is recorded and the strength is calculated. The final result is the average value of at least five valid samples.

[0023] The corrosion influence coefficient analysis module is configured to analyze the graphite heat exchanger multi-source corrosion dataset to obtain a medium corrosivity coefficient, a working condition corrosivity coefficient and a material resistance coefficient, and input the medium corrosivity coefficient, the working condition corrosivity coefficient and the material resistance coefficient into the graphite corrosion coupling prediction model to output a comprehensive corrosion index. Further, in the above technical solution, the medium corrosivity coefficient specifically comprises: , wherein C PH is a pH value of the medium, C is a chloride ion concentration, θ PH is a pH corrosion critical threshold, γ is a chloride ion concentration saturation constant, k PH is a pH sensitivity coefficient, α and β are weight factors and satisfy α+β=1, and e is a natural constant.

[0024] It should be noted that the pH corrosion critical threshold θ PH refers to a critical point at which the pH value of the medium significantly changes the corrosion behavior of the graphite material, and the value thereof is determined according to the corrosion resistance characteristics of the graphite material and long-term operation experience data, is usually obtained through laboratory accelerated corrosion test and regression analysis of historical corrosion data on site, and is used to distinguish the pH interval of high corrosion risk and low corrosion risk in the medium corrosivity coefficient calculation, when the pH value of the medium approaches or exceeds the threshold, the corrosivity coefficient will change significantly, thereby more accurately reflecting the actual corrosion trend. The chloride ion concentration saturation constant γ is used to represent the saturation effect of the chloride ion concentration on the corrosion, and the value thereof is determined through corrosion kinetics experiment and fitting analysis, reflecting the phenomenon that the corrosion rate gradually tends to be flat under high chloride ion concentration. The introduction of the constant makes the medium corrosivity coefficient not increase infinitely when the chloride ion concentration is high, which is more consistent with the nonlinear relationship between the corrosion rate and the ion concentration in actual engineering, and improves the prediction accuracy of the model under high concentration working conditions. The pH sensitivity coefficient k PH is used to quantify the sensitivity of the influence of the pH value of the medium deviating from the critical threshold on the corrosivity coefficient, and the value thereof is obtained through corrosion test data fitting and parameter optimization, which controls the slope of the change of the corrosivity coefficient with the change of the pH value. A larger k PH value indicates that a slight deviation of the pH from the threshold will cause a significant change in the corrosivity coefficient, and is suitable for high corrosion risk medium sensitive to pH change, thereby enhancing the response ability and prediction accuracy of the model to working condition fluctuations. The values of the weight factors a and b are determined according to the dominant contribution of hydrogen ions and chloride ions to the corrosion of the graphite material in the medium, and are comprehensively determined by coupling historical corrosion data, laboratory simulation test results and expert experience; specifically, a represents the weight of pH value on the corrosivity of the medium, b represents the weight of chloride ion concentration on the corrosivity of the medium, both of which satisfy the constraint condition a + b = 1, in actual application, if the corrosivity of the medium is mainly driven by acidic or alkaline conditions, such as strong acid or strong alkali working conditions, a > b is set, for example, a = 0.7, b = 0.3; if the corrosivity of the medium is mainly caused by pitting or stress corrosion caused by chloride ions, such as seawater or high-chlorine environment, b > a is set, for example, a = 0.3, b = 0.7; for mixed type corrosion medium with equivalent influence of both, balanced weight is adopted, such as a = 0.5, b = 0.5, the final value needs to be determined through repeated iteration and verification of measured data and model prediction results of corrosion rate to ensure the prediction accuracy and adaptability of the model under different working conditions.

[0025] Further, in the above technical solution, the working condition corrosivity coefficient is specifically: , Wherein, T is the medium temperature, T0 is the reference base value of the medium temperature, V is the medium flow rate, V0 is the reference base value of the medium flow rate, P is the system pressure, P0 is the reference base value of the system pressure, and η is the sensitivity adjustment index.

[0026] It should be noted that the reference base value T0 of the medium temperature is set as the rated temperature value under the design working condition of the equipment, which is determined according to the design temperature marked in the technical specification book of the graphite heat exchanger, and at the same time, the stability range of the physical property parameters of the medium at this temperature is referred to, for example, for a hydrochloric acid cooling system, T0 can be set as the operating temperature 80℃ required by the process, which needs to be lower than the critical point of thermal deformation of graphite material and have a safety margin to ensure the stability of the model calculation base when the temperature fluctuates; The reference base value V0 of the medium flow rate is the middle value of the economic flow rate in the process design, which is determined comprehensively based on fluid mechanics calculation and anti-erosion corrosion experience value, specifically, V0 is usually set as the critical flow rate of the medium in the heat exchange pipe to generate laminar flow to transition flow state, such as 0.8~1.2m / s range, which needs to be calibrated in combination with the pipe diameter, medium viscosity and graphite anti-erosion strength characteristics to avoid the synergistic effect of mechanical erosion and corrosion caused by too high flow rate; The reference base value P0 of the system pressure directly adopts the design pressure value marked on the equipment nameplate, which is derived from the maximum allowable working pressure in the strength calculation book, and when setting, 90% of the take-off pressure of the safety valve is considered as the upper threshold, for example, for PN16 pressure grade system, P0 is set as 1.6MPa, this base value can effectively distinguish the influence boundary of normal operating pressure and abnormal overpressure working condition on the corrosion behavior; The sensitivity adjustment index η is set according to the medium corrosion characteristics and system operation experience, which is used to quantify the joint influence strength of temperature, flow rate and pressure on the corrosion coefficient of the working condition. The value is obtained by regression analysis of corrosion dynamics experiment and historical operation data. Specifically, the corrosion behavior under different working condition combinations is simulated in the laboratory environment, the deviation between the actual corrosion rate and the theoretical value is measured, and the correlation between the fluctuation data of temperature, flow rate and pressure in the field operation record and the corresponding corrosion phenomenon is fitted to obtain the optimal solution range of η, which is usually 1.5~3.0. For example, for strong corrosive medium such as concentrated sulfuric acid system, η is set to a higher value of 2.5 to enhance the response ability of the model to small working condition fluctuations; for weak corrosive medium such as pure water system, it is set to a lower value of 1.8 to suppress noise interference and avoid false alarms caused by excessive sensitivity of the model.

[0027] Further, in the above technical solution, the material resistance coefficient is specifically: , Where P r is the measured graphite porosity, P r0 is the standard porosity of the new material, σ c is the measured compressive strength of graphite, σ c0 is the standard compressive strength of the new material, and λ is the amplification coefficient of porosity influence.

[0028] It should be noted that the standard porosity P r0 of the new material directly uses the measured porosity value marked in the manufacturer's quality certificate as the initial reference value for system calculation, which is a constant constant and does not change with the running state of the equipment; The standard compressive strength σ c0 of the new material is strictly set according to the measured compressive strength value specified in the manufacturer's material quality certificate. The value is measured by the compression test method specified in the national standard GB / T 13465.2-2014, using a universal material testing machine to load the standard sample of 20mm×20mm×20mm at a constant rate of 1.5mm / min to failure, and the σ c0 value is measured and marked in the factory inspection report; The setting of the amplification coefficient λ of the porosity influence is determined according to the regression analysis results of the laboratory accelerated corrosion test and the historical data, which is used to quantify the amplification degree of the non-linear acceleration effect of the increase of graphite material porosity on the corrosion rate. Specifically, the following technical solutions are realized: first, prepare a gradient porosity in the laboratory environment, such as The graphite sample with a ratio coverage in the interval of 0.8 to 1.5 is measured for a corrosion rate in a typical medium, such as 10% hydrochloric acid with a concentration of 10% and a temperature of 80°C, by using electrochemical impedance spectroscopy and weight loss method; meanwhile, the porosity growth data and the corresponding corrosion rate detection values of the heat exchanger of the same material in the same medium under different service years (1-10 years) are extracted; finally, the mapping relationship curve of the relative ratio of porosity and the corrosion rate increase is fitted by using a nonlinear least squares method, and the curve curvature characteristic parameter is extracted and converted into an optimized solution of lambda.

[0029] Further, in the technical scheme, the graphite corrosion coupling prediction model specifically comprises: , wherein Q is a comprehensive corrosion index, F is a medium corrosivity coefficient, E is a working condition aggressivity coefficient, M is a material resistance coefficient, and ω1, ω2 and ω3 are dynamic weight factors and satisfy ω1+ω2+ω3=1.

[0030] The corrosion rate prediction module is configured to perform linear mapping according to the comprehensive corrosion index in combination with a preset reference corrosion rate, calculate and output a real-time predicted corrosion rate of the graphite heat exchanger; The digital twin visualization module is configured to receive the real-time predicted corrosion rate, the comprehensive corrosion index and a multi-source corrosion data set of the graphite heat exchanger, drive a three-dimensional digital twin of the graphite heat exchanger to perform visual display, and mark a corrosion risk level. Further, in the technical scheme, the calculation formula of the real-time predicted corrosion rate is: , wherein V predicted is the real-time predicted corrosion rate, V base is the reference corrosion rate, Q is the comprehensive corrosion index, and Q base is a reference comprehensive corrosion index.

[0031] It should be noted that the setting of the reference corrosion rate V base value needs to be based on the graphite material used in the graphite heat exchanger in a new material state that meets the standard porosity P r0 and the standard compressive strength σ c0 of the new material marked in the manufacturer's quality certificate, and is determined under a reference working condition specified by the equipment design, which includes a reference reference value T0 of the medium temperature, a reference reference value V0 of the medium flow rate, a reference reference value P0 of the system pressure, and needs to match the reference medium characteristics, which are the corresponding pH corrosion critical threshold θ PHlow corrosion risk pH value and chloride ion concentration close to the low chloride benchmark value of the process design; in the determination process, the laboratory accelerated corrosion test, the measured data of the corrosion performance of the heat exchanger of the model provided by the manufacturer under the benchmark working condition, and the statistical average corrosion rate of the non-destructive testing of the long-term operation of the same type of graphite heat exchanger in the same industry under the same benchmark working condition are combined, wherein the laboratory accelerated corrosion test refers to the national standard GB / T 13465.2-2014, the standard graphite sample of the same material and the same specification as the heat exchanger is used to monitor the average corrosion rate in the preset period in the simulated corrosion environment of the benchmark working condition by the weight loss method or the electrochemical impedance spectroscopy method, and the non-destructive testing includes ultrasonic testing and penetration testing; subsequently, the above three types of data need to be subjected to data effectiveness screening and weighted average processing, the abnormal fluctuation value needs to be excluded in the data effectiveness screening, the weight of the laboratory test data is 0.4, the weight of the measured data of the manufacturer is 0.3, and the weight of the industry statistical data is 0.3, and the value of V base is obtained after processing. The setting of the benchmark comprehensive corrosion index Q base is based on the fact that the graphite heat exchanger is in a new material state, that is, the porosity and the compressive strength of the graphite material are both standard values, and the graphite heat exchanger is operated under the benchmark working condition specified by the equipment design, the benchmark working condition includes that the medium temperature, the medium flow rate and the system pressure are all reference benchmark values, at the same time, the medium pH value is in the low corrosion risk interval and the chloride ion concentration is close to the low chloride benchmark value of the process design; the value calculated by inputting the above-mentioned benchmark material parameters, benchmark working condition parameters and benchmark medium parameters into the graphite corrosion coupling prediction model is used as a normalized benchmark for corrosion prediction, and is used for proportional mapping of the real-time corrosion rate.

[0032] Further, in the above technical solution, the corrosion risk level is: When the real-time predicted corrosion rate V predicted >V1, it is a first-level corrosion risk; When the real-time predicted corrosion rate V2<V predicted ≤V1, it is a second-level corrosion risk; When the real-time predicted corrosion rate V3<V predicted ≤V2, it is a third-level corrosion risk; When the real-time predicted corrosion rate V predicted ≤V3, it is a fourth-level corrosion risk.

[0033] It should be known that V1 is a critical corrosion rate threshold, V2 is a high-medium corrosion rate dividing value, and V3 is a medium-low corrosion rate boundary value, and the values of V1, V2 and V3 are set as V1=0.08 mm / year, V2=0.15 mm / year and V3=0.25 mm / year.

[0034] The maintenance decision support module is configured to generate early warning information and maintenance suggestion work orders according to the corrosion risk level.

[0035] Further, in the above technical solution, the maintenance decision support module is further configured to perform model dynamic correction, including: a bias quantification unit configured to obtain an actual corrosion rate V actual from non-destructive testing, and calculate a relative error coefficient δ predicted according to the predicted corrosion rate V ; a correction triggering unit configured to start correction when the relative error coefficient δ th exceeds a preset threshold δ actual and the actual corrosion rate V min exceeds a minimum effective corrosion rate threshold V a weight adjustment unit configured to update the weight factor using an exponential decay method: , wherein, is a current weight factor, is an updated weight factor, μ is a learning rate, δ is a relative error coefficient, and e is a natural constant, is a weight adjustment direction function, wherein, wherein, is a partial derivative of the predicted corrosion rate with respect to the weight ω i ; a constraint processing unit configured to normalize the dynamic weight factors ω1, ω2, and ω3 and apply boundary constraints; a freeze control unit configured to suspend correction when the average relative error of the last 5 corrections is lower than a preset error threshold.

[0036] It should be noted that the initial values of the dynamic weight factors ω1, ω2, and ω3 are set as ω1=0.3, ω2=0.3, and ω3=0.4; the learning rate μ=0.2; the preset threshold δ th =0.15; the minimum effective corrosion rate threshold V min =0.1 mm / year; the boundary constraints are ; the preset error threshold is 0.05.

[0037] It needs to be known that when triggering the first level of corrosion risk, the early warning information prompts the equipment management and operation and maintenance personnel with the highest priority, and the corrosion state of the key areas such as the connection between the graphite heat exchanger tube plate and the heat exchange tube and the middle part of the heat exchange tube needs to be immediately paid attention to, so as to avoid the corrosion intensification leading to equipment failure; when triggering the second level of corrosion risk or the third level of corrosion risk, the early warning information respectively prompts the operation and maintenance personnel to strengthen the inspection frequency and data monitoring density of the corresponding area, wherein the second risk inspection frequency is increased to 2 times of the conventional, and the third risk is increased to 1.5 times; the medium pH value collection frequency is adjusted to 1 time every 30 seconds, and the chloride ion concentration collection frequency is adjusted to 1 time every 2 minutes, so as to track the corrosion trend change in real time; when triggering the fourth level of corrosion risk, the early warning information prompts that the equipment is in a low corrosion risk state, and the management can be carried out according to the conventional monitoring period, such as PH sensor calibration every month, chloride ion concentration sensor calibration every quarter, and at the same time, all level early warning information needs to be clearly marked with the current comprehensive corrosion index and each influence coefficient, including the medium corrosivity coefficient, the working condition corrosivity coefficient and the material resistance coefficient, so as to accurately locate the corrosion dominant factor for the operation and maintenance personnel; The maintenance suggestion work order is generated by the maintenance decision support module according to the corrosion risk level, the real-time predicted corrosion rate and the non-destructive testing data, and the work order needs to clearly indicate the maintenance target, specific measures, implementation period, detection means and data feedback requirements; for the first level of corrosion risk, the work order suggests stopping immediately, using the straight probe ultrasonic detection method with a frequency of 2.5-5 MHz to comprehensively detect the corrosion of key structural parts such as heat exchange tubes, tube plates and shells, and combining with the penetration detection to confirm the local corrosion condition, and after detection, the body state data needs to be updated according to the measured porosity and compressive strength, if the corrosion exceeds the standard, a component replacement scheme is developed, and the replaced component needs to meet the manufacturer's quality standard, such as the new material porosity P r0 and the compressive strength σ c0 , after repair, multi-source corrosion data needs to be re-collected to verify the maintenance effect; for the second level of corrosion risk, the work order suggests shortening the regular maintenance period by 50%, and performing ultrasonic detection on the key area once a week, increasing the collection frequency of medium temperature, flow rate and pressure, such as collecting temperature every 15 seconds, collecting pressure every 30 seconds, and additionally calibrating the PH sensor and the chloride ion concentration sensor once a month to avoid data deviation affecting corrosion prediction; for the third level of corrosion risk, the work order suggests performing maintenance according to the regular maintenance period, performing key area ultrasonic detection once every two weeks, and maintaining the regular calibration frequency of the sensor, i.e. PH once a month and chloride ion once a quarter, and at the same time, regularly cleaning the sensor probe dirt, for example, the chloride ion sensor is cleaned once every two months; for the fourth level of corrosion risk, the work order suggests maintaining according to the original design maintenance period, and performing regular inspection and sensor calibration.

[0038] The application provides a graphite heat exchanger corrosion rate prediction method based on digital twinning as shown in the formula I: Figure 2 , and specifically comprises the following steps: S1, the multi-source corrosion data collection module collects the medium characteristics, working condition operation parameters and body state data of the graphite heat exchanger through a plurality of sensors deployed on the graphite heat exchanger body and upstream and downstream process pipelines and through non-destructive testing means during the maintenance period, to form a multi-source corrosion data set of the graphite heat exchanger; S2, the corrosion influence coefficient analysis module is used for analyzing the multi-source corrosion data set of the graphite heat exchanger to obtain a medium corrosivity coefficient, a working condition corrosivity coefficient and a material resistance coefficient, and inputting the medium corrosivity coefficient, the working condition corrosivity coefficient and the material resistance coefficient into a graphite corrosion coupling prediction model to output a comprehensive corrosion index; S3, the corrosion rate prediction module is used for linear mapping according to the comprehensive corrosion index in combination with a preset reference corrosion rate, calculating and outputting a real-time predicted corrosion rate of the graphite heat exchanger; S4, the digital twin visualization module is used for receiving the real-time predicted corrosion rate, the comprehensive corrosion index and the multi-source corrosion data set of the graphite heat exchanger, driving a three-dimensional digital twin of the graphite heat exchanger to perform visual display, and marking a corrosion risk level; S5, the maintenance decision support module is used for generating early warning information and maintenance suggestion work orders according to the corrosion risk level.

[0039] Finally, it should be noted that: the above only for the preferred embodiments of the present application, and not for limiting the present application, although the foregoing detailed description of the present application is made with reference to the foregoing embodiments, for the skilled in the art, it still can be modified, or part of the technical features of the equivalent replacement, within the spirit and principles of the present application, any modification, equivalent replacement, improvement, etc., should be included in the scope of protection of the present application.

Claims

1. A graphite heat exchanger corrosion rate prediction system based on digital twinning, characterized in that, The method comprises the following steps: A multi-source corrosion data acquisition module acquires medium characteristics, operating parameter and body state data of the graphite heat exchanger through a plurality of sensors arranged on the graphite heat exchanger body and upstream and downstream process pipelines and through nondestructive testing means during a maintenance period, to form a multi-source corrosion data set of the graphite heat exchanger; An corrosion influence coefficient analysis module is configured to analyze the multi-source corrosion data set of the graphite heat exchanger to obtain a medium corrosivity coefficient, an operating corrosivity coefficient and a material resistance coefficient, and input the medium corrosivity coefficient, the operating corrosivity coefficient and the material resistance coefficient into a graphite corrosion coupling prediction model to output a comprehensive corrosion index; A corrosion rate prediction module is configured to calculate and output a real-time predicted corrosion rate of the graphite heat exchanger according to the comprehensive corrosion index and a preset reference corrosion rate through linear mapping; A digital twin visualization module is configured to receive the real-time predicted corrosion rate, the comprehensive corrosion index and the multi-source corrosion data set of the graphite heat exchanger, drive a three-dimensional digital twin of the graphite heat exchanger to perform visual display, and mark a corrosion risk level; A maintenance decision support module is configured to generate early warning information and maintenance suggestion work orders according to the corrosion risk level.

2. The graphite heat exchanger corrosion rate prediction system based on digital twinning of claim 1, wherein, The graphite heat exchanger multi-source corrosion data set includes medium characteristic data, working condition operation data and body state data; the medium characteristic data includes medium PH value C PH , chloride ion concentration C; the working condition operation data includes medium temperature T, medium flow rate V, system pressure P; and the body state data includes graphite material porosity P r , graphite material compressive strength σ c .

3. The graphite heat exchanger corrosion rate prediction system based on digital twinning of claim 1, wherein, The medium corrosivity coefficient is specifically: ， where C PH is the medium pH value, C is the chloride ion concentration, θ PH is the pH corrosion critical threshold, γ is the chloride ion concentration saturation constant, k PH is the pH sensitivity coefficient, α and β are weight factors and satisfy α + β = 1, and e is the natural constant.

4. The digital-twin-based graphite heat exchanger corrosion rate prediction system of claim 1, wherein, The operating corrosivity coefficient is specifically: ， wherein T is a medium temperature, T0 is a reference reference value of the medium temperature, V is a medium flow rate, V0 is a reference reference value of the medium flow rate, P is a system pressure, P0 is a reference reference value of the system pressure, and η is a sensitivity adjustment index.

5. The digital-twin-based graphite heat exchanger corrosion rate prediction system of claim 1, wherein, The material resistance coefficient is specifically: ， where P r is the measured porosity of the graphite, P r0 is the standard porosity of the new material, σ c is the measured compressive strength of the graphite, σ c0 is the standard compressive strength of the new material, and λ is a porosity-affected amplification factor.

6. The digital-twin-based graphite heat exchanger corrosion rate prediction system of claim 1, wherein, The graphite corrosion coupling prediction model is specifically: ， wherein Q is the comprehensive corrosion index, F is the medium corrosivity coefficient, E is the operating corrosivity coefficient, M is the material resistance coefficient, ω1, ω2 and ω3 are dynamic weight factors and satisfy ω1+ω2+ω3=1.

7. The digital-twin-based graphite heat exchanger corrosion rate prediction system of claim 1, wherein, The calculation formula of the real-time predicted corrosion rate is: ， wherein V predicted is the real-time prediction of the corrosion rate, V base is the baseline corrosion rate, Q is the integrated corrosion index, Q base is the baseline integrated corrosion index.

8. The digital-twin-based graphite heat exchanger corrosion rate prediction system of claim 1, wherein, The corrosion risk level is: When the real-time predicted corrosion rate V predicted V1, it is a first-class corrosion risk; When the real-time predicted corrosion rate V2 < V predicted ≤ V1, it is a secondary corrosion risk; When the real-time predicted corrosion rate V3 < V predicted ≤V2, it is a third-level corrosion risk; When the real-time predicted corrosion rate V predicted ≤ V3, it is a fourth-grade corrosion risk.

9. The digital-twin-based graphite heat exchanger corrosion rate prediction system of claim 1, wherein, The maintenance decision support module is further configured to perform model dynamic correction, comprising: a bias quantification unit configured to obtain an actual corrosion rate V actual with the real-time predicted corrosion rate V predicted a relative error coefficient δ is calculated: ； a correction trigger unit for initiating a correction when the relative error coefficient δ exceeds a preset threshold value δ th and the actual corrosion rate V actual exceeds a minimum effective corrosion rate threshold value V min value A weight adjustment unit updates the weight factors by using an exponential decay method: ， wherein, is the current weight factor, is the updated weight factor, μ is the learning rate, δ is the relative error coefficient, e is the natural constant, is the weight adjustment direction function, wherein, wherein, is the partial derivative of the predicted corrosion rate with respect to the weight ω i . A constraint processing unit is configured to normalize the dynamic weight factors ω1, ω2 and ω3 and apply boundary constraints; A freezing control unit is configured to suspend correction when the average relative error of the last five corrections is lower than a preset error threshold.

10. A method for predicting the corrosion rate of a graphite heat exchanger based on digital twinning, characterized by, The method comprises the following steps: S1, a multi-source corrosion data acquisition module acquires medium characteristics, operating parameter and body state data of the graphite heat exchanger through a plurality of sensors arranged on the graphite heat exchanger body and upstream and downstream process pipelines and through nondestructive testing means during a maintenance period, to form a multi-source corrosion data set of the graphite heat exchanger; S2, an corrosion influence coefficient analysis module is configured to analyze the multi-source corrosion data set of the graphite heat exchanger to obtain a medium corrosivity coefficient, an operating corrosivity coefficient and a material resistance coefficient, and input the medium corrosivity coefficient, the operating corrosivity coefficient and the material resistance coefficient into a graphite corrosion coupling prediction model to output a comprehensive corrosion index; S3, a corrosion rate prediction module is configured to calculate and output a real-time predicted corrosion rate of the graphite heat exchanger according to the comprehensive corrosion index and a preset reference corrosion rate through linear mapping; S4, the digital twin visualization module is configured to receive the real-time predicted corrosion rate, the comprehensive corrosion index, and the multi-source corrosion data set of the graphite heat exchanger, drive a three-dimensional digital twin of the graphite heat exchanger to perform visual display, and mark a corrosion risk level; S5, the maintenance decision support module is configured to generate early warning information and a maintenance suggestion work order according to the corrosion risk level.

Images