Liquid flow battery electrolyte activity recovery method

Abstract

This application discloses a method for restoring the activity of an electrolyte in a flow battery. The method includes: acquiring multimodal data streams of the flow battery under charge-discharge operation; evaluating the electrolyte activity state based on the multimodal data streams and calculating the capacity decay rate; and when the capacity decay rate exceeds a first threshold, performing a first-level operating condition recovery to forcibly adjust operating parameters to suppress operating condition-induced activity decay. In step S1, this invention simultaneously acquires electrochemical data streams, physical property data streams, and side reaction characteristic data streams, and in steps S2 to S4, it collaboratively utilizes these data for phased judgment. Compared to the threshold triggering strategy in the prior art that relies solely on a single indicator such as capacity retention rate or voltage, this invention can identify potential risks in the early stages of capacity decay through abnormal increases in viscosity or growth trends in charge transfer resistance, achieving earlier and more accurate decay warnings.

Description

Technical Field

[0001] This application belongs to the field of flow battery technology, specifically relating to a method for restoring the activity of electrolytes in flow batteries. Background Technology

[0002] Currently, the flow battery field has developed model-based State of Health (SOH) estimation and threshold-based maintenance strategies, which can predict capacity decay trends to some extent and trigger interventions. For example, this can be achieved by identifying changes in ohmic internal resistance or charge transfer resistance online, or by setting a rebalancing operation when the capacity retention rate falls below a certain fixed threshold (e.g., 80%). However, existing intervention strategies still have the following shortcomings:

[0003] (1) Single trigger dimension: Most strategies rely on a single electrochemical index (such as voltage, capacity retention rate) or a fixed time / cycle period, and fail to fully integrate multimodal information such as electrochemical data (such as real-time impedance spectrum), physical property data (such as electrolyte viscosity, conductivity) and side reaction characteristic data (such as hydrogen evolution rate) for comprehensive judgment.

[0004] (2) Insufficient ability to identify the type of capacity decay: Capacity decay may be caused by operating condition-induced factors (such as excessively wide SOC range, unsuitable current density, osmotic pressure imbalance), or by electrolyte aging (such as deactivation of active complexes, depletion of additives) or passivation of electrode interfaces. Existing strategies are unable to distinguish the above types of decay in the early stage of decay (when the capacity retention rate is still higher than 95%), resulting in a lack of targeted selection of intervention strategies.

[0005] (3) It is easy to cause over-repair or under-repair: For condition-induced degradation, if chemical addition or forced repair is used directly, it will cause unnecessary additive consumption and system downtime (over-repair); while for body aging or interface passivation, adjusting the operating conditions alone will not be effective (under-repair), delaying the best intervention time.

[0006] Therefore, there is an urgent need for an electrolyte activity recovery method that can integrate multimodal data, identify the type of decay in stages, and implement progressive interventions to achieve precise maintenance that "solves the right problem at the right time with the right means". Summary of the Invention

[0007] This application provides a method for restoring the electrolyte activity of a flow battery, aiming to solve the problems of delayed electrolyte activity restoration, isolated restoration methods, lack of interfacial activity assessment, and poor adaptability to different decay modes in the prior art.

[0008] A method for restoring electrolyte activity in a flow battery, the method comprising:

[0009] Collect multimodal data streams of flow batteries during charge and discharge operation. The multimodal data streams include at least electrochemical data streams, physical characteristic data streams, and capacity decay characteristic data streams, and synchronize and align the multimodal data streams according to a unified timestamp.

[0010] The electrolyte activity state is evaluated based on the multimodal data stream, and the capacity decay rate is calculated. When the capacity decay rate exceeds a first threshold, the first-level operating condition recovery is executed, and the operating parameters are forcibly adjusted to suppress operating condition-induced activity decay.

[0011] If the capacity decay rate still exceeds the first threshold after the first-level operating condition recovery is performed, the physicochemical properties of the electrolyte are tested, and composite additives are injected according to the test results to perform the second-level chemical addition recovery to compensate for the electrolyte's intrinsic activity.

[0012] If the capacity decay rate still exceeds the first threshold after performing the second-level chemical addition recovery, the electrochemical impedance spectroscopy is measured to extract the charge transfer resistance. When the charge transfer resistance exceeds a preset multiple of the initial charge transfer resistance, the third-level forced electrochemical repair is performed to restore the activity of the electrode / electrolyte interface.

[0013] Optionally, the multimodal data stream further includes a side reaction feature data stream, which includes at least a cumulative hydrogen evolution rate. When the cumulative hydrogen evolution rate continuously exceeds the hydrogen evolution threshold, a hydrogen evolution anomaly flag is generated to trigger the short-time overpotential repair protocol in the third-level forced electrochemical repair.

[0014] Optionally, the calculation of the capacity decay rate includes: extracting the capacity retention rate sequence of the most recent N complete charge-discharge cycles from the synchronized multimodal data stream, performing linear fitting on the capacity retention rate sequence using the least squares method, and using the slope obtained from the fitting as the current capacity decay rate; the first threshold value ranges from 0.02% / cycle to 0.08% / cycle.

[0015] Optionally, the first-level operating condition recovery includes: forcibly clamping the charge / discharge cutoff SOC between 0.4 and 0.6, and switching the current density from a first current density to a second current density when the SOC reaches 0.6 during charging, wherein the first current density ranges from 180 to 220 mA / cm². 2 The second current density ranges from 140 to 180 mA / cm². 2 .

[0016] Optionally, the physicochemical performance testing of the electrolyte includes: detecting the valence concentration of vanadium ions or iron / chromium ions in the electrolyte by ultraviolet-visible spectrophotometry, and / or detecting the relative proportion of active complexes in the chromium-based flow battery by Raman spectroscopy; when the relative proportion of the active complexes is less than 60%, the electrolyte is determined to be aged.

[0017] Optionally, the composite additive includes inorganic and organic additives, wherein the inorganic additive is selected from ferric ammonium sulfate or antimony trichloride, and the organic additive is selected from glycerol; for vanadium redox flow batteries, ferric ammonium sulfate is injected to achieve a final concentration of 0.3%–0.5% in the electrolyte; for iron-chromium redox flow batteries, glycerol is injected to achieve a final concentration of 0.5%–0.8%, while antimony trichloride is injected simultaneously. The final ion concentration reached 5–15 mmol / L.

[0018] Optionally, the measurement of the electrochemical impedance spectroscopy includes: applying a sinusoidal AC excitation signal with an amplitude of 5 mV and a frequency range of 100 kHz to 0.01 Hz to the fuel cell stack, using an equivalent circuit model. The fitting process is performed to extract the ohmic resistance, charge transfer resistance, and constant phase angle element index; the preset multiple is 2.0 times.

[0019] Optionally, the third-level forced electrochemical repair includes an asymmetric pulse repair protocol: applying a sequence of positive and negative asymmetric high-frequency pulse voltages to the stack, with a pulse frequency of 100–500 Hz, a positive pulse amplitude of 1.2 times the rated charging voltage, a negative pulse amplitude of 0.8 times the rated charging voltage, a positive pulse width to negative pulse width ratio of 1:5, and a total repair time of 30–60 seconds, until the charge transfer resistance drops to less than 1.5 times the initial charge transfer resistance.

[0020] Optionally, the third-level forced electrochemical repair includes a short-time overpotential repair protocol: when the battery system is an iron-chromium flow battery and the cumulative hydrogen evolution rate continuously exceeds 0.1 mL / min, the hydrogen gas evolved at the top of the negative electrode tank is introduced into the anode chamber of the hydrogen-iron rebalancing unit, enriching the positive electrode tank with hydrogen gas that has been generated due to hydrogen evolution. The electrolyte is introduced into the cathode chamber of the rebalancing unit and run for 15–30 minutes until it reaches the positive electrode electrolyte. The concentration has dropped to the normal range.

[0021] Optionally, the method further includes final recovery verification: after performing the third-level forced electrochemical repair, the capacity decay rate after repair and the discharge capacity of the latest cycle are recalculated. If the capacity decay rate after repair does not exceed the first threshold and the discharge capacity of the latest cycle is not less than 0.85 times the rated capacity, the recovery is determined to be successful and the system is switched back to the optimization mode; otherwise, a maintenance alarm signal is issued and the battery system is placed in a safe standby state.

[0022] Compared with the prior art, this application has at least the following beneficial effects:

[0023] In step S1, this invention simultaneously acquires electrochemical data streams, physical property data streams, and side reaction characteristic data streams. In steps S2 to S4, the above data are used in a coordinated manner to make phased judgments. Compared with the threshold triggering strategy in the prior art that only relies on a single indicator such as capacity retention rate or voltage, this invention can identify potential risks in the early stage of capacity decay by abnormal increase in viscosity or increase in charge transfer resistance, thus achieving earlier and more accurate decay warning.

[0024] This invention employs a three-tiered progressive judgment mechanism to distinguish three different causes of activity degradation: condition-induced degradation, electrolyte component imbalance or aging, and electrode / electrolyte interface passivation or side reaction accumulation. For each type of degradation, differentiated intervention methods are adopted, such as condition optimization, composite additive compensation, and forced electrochemical repair. Compared with the existing technology's strategy of "uniform threshold triggering and single intervention method," this invention avoids misjudging condition-induced degradation as bulk aging and thus performing unnecessary chemical additions, and also avoids delaying intervention opportunities by only adjusting the operating conditions for bulk aging. This significantly improves the pertinence and effectiveness of recovery measures.

[0025] This invention triggers the process step by step in the order of "operating condition optimization → chemical addition → forced repair," and only enters the next level when the previous level is insufficient. This "minimum necessary intervention" principle minimizes additive consumption, system downtime, and electrode material loss. Attached Figure Description

[0026] Figure 1 This is a flowchart of a method for restoring the electrolyte activity of a flow battery according to an embodiment of this application. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments.

[0028] The method for restoring electrolyte activity in a flow battery provided in this application, such as Figure 1 As shown, it includes the following steps:

[0029] S1: Real-time acquisition and feature extraction of multimodal operating data. While the flow battery (including vanadium redox flow batteries, iron-chromium flow batteries, and their hybrid systems) is in charge-discharge operation, four types of multimodal data streams are continuously acquired at a variable sampling period of 1-10 seconds (preferably 2 seconds or 5 seconds, dynamically adjusted according to battery capacity and operating conditions). This data is then transmitted in real-time to the central controller (an embedded industrial computer or PLC with a built-in data buffer and sliding time window module) via industrial Ethernet or CAN bus. The four types of multimodal data streams specifically include:

[0030] Electrochemical data streams are continuously recorded using high-precision voltage sensors (accuracy ±0.1%, range 0–2V) connected in parallel to the positive and negative terminals of the battery stack, and a Hall current sensor (accuracy ±0.2%) connected in series in the main circuit, to continuously record real-time voltage V(t) and current density J(t). Simultaneously, an online SOC monitor (based on open-circuit voltage correction and coulomb integral coupling algorithm) integrated into the positive and negative electrolyte circulation pipelines is used to obtain the state of charge of the positive electrolyte. State of charge of the negative electrode electrolyte The SOC data is updated every 5 seconds and automatically compensates for electromotive force shifts caused by temperature changes.

[0031] Physical property data streams were generated by installing online vibratory viscometers (measuring range 0–50 mPa·s, accuracy ±1.0%) and conductivity probes (electrode constant 1.0, frequency 1 kHz) in the outlet pipelines of the positive and negative electrolyte storage tanks, respectively, to continuously measure the real-time viscosity of the positive electrolyte. Real-time viscosity of negative electrode electrolyte and the conductivity of the mixed electrolyte To prevent bubble interference, a miniature degassing device is installed at the front end of the viscometer, and the median of three consecutive measurements is taken and filtered before being uploaded at each sampling point (every 10 seconds).

[0032] The capacity decay characteristic data stream is used by the central controller to calculate the discharge capacity of each complete charge-discharge cycle (from charging cutoff to discharging cutoff) online using the ampere-hour integration method based on the collected real-time voltage V(t) and current density J(t). Where n is the cycle number. Furthermore, by calling the pre-stored rated capacity in memory... (This rated capacity is the average of the first three cycles of the battery system under standard operating conditions.) Calculate the capacity retention rate for the current cycle. To avoid the influence of measurement noise, the controller performs five consecutive cycles. Perform Savitzky-Golay smoothing filtering and store the filtered sequence in a decay trend queue;

[0033] Side reaction characteristic data stream (applicable to iron-chromium flow batteries): A piezoresistive gas pressure sensor (range 0–10 kPa, accuracy ±0.5%) is installed in the sealed space at the top of the negative electrode electrolyte storage tank to monitor the gas phase pressure inside the tank in real time. The central controller calculates the cumulative hydrogen evolution rate every 30 seconds based on the ideal gas law, combined with the tank headspace volume and electrolyte temperature. ,in Let R be the headspace volume of the tank, and R be the gas constant. This is the real-time temperature. When When the sampling rate exceeds 0.05 mL / min for three consecutive sampling cycles, the controller automatically marks the hydrogen evolution abnormality flag and uses it as one of the important criteria for triggering the forced repair protocol in S4;

[0034] All the aforementioned data streams are synchronized and aligned according to a unified timestamp within the central controller and stored in a circular buffer for use by S2 in real-time decay rate calculation and first-level recovery trigger judgment. If a data stream (such as online viscosity value) exceeds the reasonable range three times consecutively (e.g., η>20mPa·s or η<0.5mPa·s), the controller will temporarily block the data source, perform interpolation based on historical trends, and issue a sensor self-check alarm.

[0035] S2: Electrolyte activity status assessment and first-level operating condition recovery. The central controller performs real-time analysis of the multimodal data stream collected and synchronized in S1, and executes the first-level recovery judgment and operating condition intervention according to the following steps:

[0036] S2.1: Calculate the capacity decay rate. The controller extracts the capacity retention rate sequence of the most recent N complete charge-discharge cycles from the circular buffer. Where N≥10, preferably N=15. The least squares method is used to perform linear fitting on the sequence, and the fitting model is: The slope 'a' represents the current capacity decay rate. The unit is % / cycle. To ensure the reliability of the fit, the correlation coefficient is required to be [value missing]. ;like If so, expand the window to N=20 and refit until the correlation requirement is met. The calculated... It is sent to the first-level recovery trigger judgment module;

[0037] S2.2: First-level recovery trigger and execution, the controller will Compared with the first threshold pre-stored in non-volatile memory The comparison. The value range is 0.02% / cycle to 0.08% / cycle, and in this embodiment, it is preferably 0.05% / cycle. When If this condition is met for two consecutive cycles, it is determined to be a condition-induced activity decay, and the first-level condition recovery is immediately triggered.

[0038] Upon triggering, the central controller sends an "optimization mode" command to the battery management system (BMS) or charging / discharging equipment (bidirectional DC-DC converter) via the digital output module, forcibly adjusting the following operating parameters:

[0039] SOC operating range limitation: The charge / discharge cutoff SOC is forcibly clamped between 0.4 and 0.6. That is, charging stops when the SOC of the electrolyte at any electrode reaches 0.6, and discharging stops when the SOC drops to 0.4. This range is determined based on a built-in historical data model—this model, through offline testing of vanadium ion transmembrane migration at different SOC ranges, confirms that the net vanadium ion migration per unit cycle is lowest in the 0.4–0.6 range (measured values ​​are lower than...). (mol), with the lowest concentration polarization overpotential;

[0040] Variable current strategy: During charging, when the SOC is detected to reach 0.6 (note: this 0.6 is the variable current switching point, not the charging cutoff point), the controller sends a current density switching command to the charger, changing the current density from the current higher value. Switch to a lower value .in The range of values ​​is , The range of values ​​is In this embodiment, the preferred option is... , Keep after switching Continue charging until the SOC reaches 0.6, then stop charging.

[0041] The discharge process is similar; the current density is reduced from [previous value] when the SOC drops to 0.4. Switch to This variable current strategy extends the discharge time and increases the depth of discharge. By reducing the final reaction current density, it effectively suppresses the formation of a depletion layer on the electrode surface, thereby mitigating the loss of usable capacity caused by concentration polarization.

[0042] S2.3: Recovery effect monitoring. After the above operating condition adjustment is performed, the system returns to S1 to continue collecting data in real time and continues to execute S2.1 to calculate the new capacity decay rate. A monitoring window length M is set, where M ≥ 5, and in this embodiment, M = 10 complete cycles is preferred. If the monitoring window ends... If the first-level recovery is successful, the system will continue to run in "optimization mode", write the recovery success event to the log, and reset the continuous insufficient recovery counter.

[0043] Conversely, if the monitoring window ends If this occurs, it is determined that the first-level recovery is insufficient. At this time, the controller will set the status flag to "". "Set to 1 and automatically switch to S3 to trigger the second-stage chemical addition recovery process. At the same time, the controller packages the capacity decay rate change curve and operating condition adjustment records during the first-stage recovery period into a data package for reference when calculating the dosage of composite additives in S3;

[0044] S3: Electrolyte physicochemical performance evaluation and second-stage chemical addition recovery. If, in S2, it is determined that the first-stage operating condition recovery is insufficient (i.e., the status flag is "..."),... When "S1" is true, the central controller automatically initiates the second-level recovery process, which evaluates the physicochemical properties of the electrolyte offline or online, and implements precise compensation with composite additives based on the evaluation results. Specifically, S3 includes the following steps:

[0045] S3.1: Offline / Online physicochemical property detection. The controller sends a sampling command to the automatic sampling device integrated in the electrolyte circulation pipeline. The automatic sampling device includes a two-position three-way solenoid valve, a micro diaphragm pump, and a quantitative sampling loop, with a sampling volume of 2-5 mL per sample. The sampling device extracts electrolyte samples from the bottom outlet pipelines of the positive and negative electrode tanks respectively and delivers them to the integrated analysis module. This analysis module includes the following detection units:

[0046] Ultraviolet-Visible Spectrophotometric Detection Unit: A dual-beam ultraviolet-visible spectrophotometer (wavelength range 200–1100 nm, resolution 1 nm) is used, with deionized water as a reference, to scan the absorption spectrum of the electrolyte in the 400–800 nm wavelength range. For vanadium redox flow batteries, based on tetravalent vanadium ions (… The characteristic absorption peak intensity at approximately 766 nm and the pentavalent vanadium ion ( The shoulder peak at approximately 630 nm was compared with a pre-established standard curve to calculate the concentration of the positive electrode electrolyte. and The actual concentration. For iron-chromium flow batteries, it depends on... At approximately 340nm and Quantitative analysis was performed using the characteristic absorption peak at approximately 570 nm. Each sample was measured three times, and the average value was used as the final concentration data.

[0047] Raman spectroscopy detection unit: Employs a confocal micro Raman spectrometer (excitation wavelength 532 nm, laser power 10 mW, spectral resolution 2). The electrolyte sample was injected into a quartz cuvette and heated at 200–2000 °C. Raman spectra were collected within the specified range. For iron-chromium flow batteries, the focus was on approximately 520 nm. Agreement 400 The characteristic peaks at these locations correspond to the active complexes, respectively. Inactive complexes The controller calculates the relative proportion of active complexes using the peak area integration method. ,in for peak area, for The peak area. If If so, it is determined that the electrolyte exhibits significant aging.

[0048] The above detection results (including the concentration of each valence ion and the proportion of active complexes) along with the decay curve data recorded in S2 during the first-level recovery period are packaged and sent to the dose calculation module of the central controller.

[0049] S3.2: Precise compensation of compound additives. The dosage calculation module of the central controller determines the type and dosage of the required additives based on the detection results of S3.1 and the built-in rule library, and controls the micro-injection pump to perform the injection operation.

[0050] The composite additive consists of inorganic and organic additives, which are pre-prepared as standard concentration mother liquors and stored in additive storage tanks. The inorganic additives are selected from ferric ammonium sulfate (… ) or antimony trichloride ( The organic additives are selected from glycerol ( In this embodiment, the concentration of ferric ammonium sulfate mother liquor is 0.5 mol / L, the concentration of antimony trichloride mother liquor is 0.1 mol / L, and the glycerol mother liquor is a 10% (v / v) aqueous solution;

[0051] Dosage calculations follow these rules:

[0052] For all-vanadium redox flow batteries: when detected and When the concentration ratio deviates from 1:1 by more than 20%, it is considered a vanadium ion concentration imbalance. In this case, the required amount of ferric ammonium sulfate to be injected is calculated to achieve a final concentration of 0.3%–0.5% (volume percentage, based on the mother liquor) in the electrolyte. The addition of ferric ammonium sulfate can improve the conductivity of the electrolyte and reduce the charge transfer resistance, thereby improving electrode reaction kinetics.

[0053] For iron-chromium redox flow batteries: when the proportion of active complexes is detected. At this point, it is determined that the chromium ion activity has decreased. The required amount of composite additive to be injected is then calculated to ensure that the final concentration of glycerol in the electrolyte reaches 0.5%–0.8% (volume percentage), while simultaneously ensuring… The final concentration of ions reaches 5–15 mmol / L (preferably 15 mmol / L). Glycerol provides additional reactive sites through its hydroxyl groups, promoting… Electron transfer in a redox couple; During the charging and discharging process, ions are electrodeposited as metallic antimony particles on the surface of the graphite felt electrode. These particles... The reaction has a catalytic effect, and at the same time, its high hydrogen evolution overpotential effectively inhibits side reactions;

[0054] The injection operation is controlled by a central controller via an RS485 bus using a micro-injection pump (accuracy ±1%, flow rate range 0.1~10mL / min). The additive is injected in multiple doses, each dose being 20% ​​of the calculated total dose. After each injection, the system continues to run for 2~3 cycles to ensure the additive is fully mixed and effective, avoiding excessive local concentrations or precipitation caused by a single injection.

[0055] S3.3: Recovery effect monitoring. After all additives are injected, the system returns to S1 and S2 to continue data acquisition and capacity decay rate calculation. A monitoring window length L is set, L≥3, and in this embodiment, L=5 complete charge-discharge cycles is preferred. If the monitoring window ends... If the first threshold is reached, the second-level recovery is considered successful. The system continues to operate while maintaining the current electrolyte state, and the successful recovery event is written to the long-term log. At the same time, the recovery failure flags at each level are reset.

[0056] Conversely, if the monitoring window ends If this occurs, it is determined to be a deep-level activity decay, indicating insufficient recovery at the second stage. At this point, the controller will set the status flag to "". "Set to 1 and automatically switch to S4, triggering the third-level forced electrochemical repair process. At the same time, the controller packages and stores the current physicochemical test data, additive injection records, and subsequent decay rate evolution curves for use in S4 when optimizing the repair protocol parameters;

[0057] S4: Electrochemical interface assessment and third-level forced repair and recovery. When S3 determines that the second-level chemical addition recovery is insufficient (i.e., the status flag is " "), the second-level chemical addition recovery is insufficient (i.e., the status flag is " "). When the value is 1), the central controller determines that the chemical composition of the electrolyte can no longer be effectively corrected by additives, and the degradation is mainly due to passivation or byproduct accumulation at the electrode / electrolyte interface. At this point, the third-level forced repair process is automatically triggered to restore interfacial activity through in-situ electrochemical assessment and active repair protocols. Specifically, S4 includes the following steps:

[0058] S4.1: Online electrochemical impedance spectroscopy measurement. When the battery system is in a quiescent state (charge / discharge current is zero) or during the charge / discharge transition interval (typically within a 30-second waiting window after each change in charge / discharge direction), the central controller starts the electrochemical workstation via a digital interface, applying a sinusoidal AC excitation signal to the battery stack. The excitation signal is a sine wave with an amplitude of 5mV (peak-to-peak value, relative to open-circuit potential), and the scanning frequency range is from a high frequency of 100kHz to a low frequency of 0.01Hz. At least 40 frequency points are taken at logarithmic intervals, and each frequency point is integrated for 3 cycles to reduce noise.

[0059] The measured impedance spectrum data is transmitted to the controller in real time. The controller uses ZView or equivalent circuit fitting software to model the equivalent circuit. Perform nonlinear least squares fitting and extract the following parameters:

[0060] Ohm resistor The intersection of the impedance spectrum with the real axis in the high-frequency region (>10kHz) reflects the sum of the bulk resistance of the electrolyte, the membrane resistance, and the bulk resistance of the electrode material.

[0061] Charge transfer resistance The diameter of the semicircle corresponding to the mid-frequency region reflects the resistance required for electrons to cross the electrode / electrolyte interface during the electrochemical reaction.

[0062] CPE index of constant phase angle element: reflects the change in electrode surface roughness or porosity;

[0063] The controller will measure the data this time. The initial charge transfer resistance stored during the initial operation phase of the battery (average of the first 10 cycles) Compare. If If the condition is met in two consecutive independent measurements (with an interval of no less than 10 cycles), it is determined that there is severe passivation, adsorption of inert substances, or accumulation of by-product crystals on the electrode surface, and forced repair is required.

[0064] S4.2: Forced electrochemical repair protocol execution. After the above triggering conditions are met, the central controller first sends a "pause normal charging and discharging" command to the battery management system, disconnects the main circuit contactor, and isolates the battery stack from the charging and discharging equipment. Subsequently, based on the battery type and the side reaction characteristic data stream recorded in S1, one or more of the following repair protocols are selected and executed:

[0065] Protocol A: Asymmetric Pulse Repair (applicable to general interface repair of vanadium redox flow batteries and iron-chromium flow batteries). The controller applies a sequence of positive and negative asymmetric high-frequency pulse voltages to the battery stack through a dedicated pulse generation module (with built-in IGBT full-bridge circuit). The pulse parameters are set as follows: pulse frequency is 100–500 Hz, preferably 200 Hz in this embodiment; the positive pulse amplitude is 1.2 times the rated charging voltage of the battery system (e.g., for a single cell with a rated voltage of 1.4V, the positive pulse amplitude is 1.68V), and the pulse width is 1 ms; the negative pulse amplitude is 0.8 times the rated charging voltage (1.12V), and the pulse width is 5 ms. In each pulse cycle (6 ms), the duty cycle of the positive pulse is approximately 16.7%, and the duty cycle of the negative pulse is approximately 83.3%. The total repair time is 30–60 seconds, which can be adjusted according to… Real-time monitoring value dynamic adjustment: After applying a 10-second pulse, pause for 2 seconds and measure once. ,when Drop to below Repairs may be terminated prematurely.

[0066] The mechanism of this pulse sequence is as follows: the high anodic overpotential generated by the positive pulse promotes the electro-oxidative desorption of inactive organic matter (such as aged complexes or additive residues) adsorbed on the electrode surface; the high cathodic overpotential generated by the negative pulse can reduce part of the metal oxide passivation layer, while the short-term local strong electric field helps to destroy the physical adhesion of crystallization byproducts. The asymmetric design of the positive and negative pulses (the negative pulse is longer) avoids irreversible oxidative damage to the electrode material caused by prolonged high potential.

[0067] Protocol B: Short-term overpotential repair (specifically for hydrogen evolution imbalance repair in iron-chromium flow batteries), if the cumulative hydrogen evolution rate recorded in S1 If the sampling rate exceeds 0.1 mL / min for three consecutive sampling cycles, and the battery system is an iron-chromium redox flow battery, this protocol will be executed first. The controller instructs the pneumatic valve to introduce hydrogen gas (approximately 95%–99% by volume, the remainder being water vapor and trace amounts of nitrogen) evolved from the top of the negative electrode tank into the anode chamber of the auxiliary equipment—the "hydrogen-iron rebalancing unit"—through a stainless steel gas pipe. Simultaneously, a portion of the hydrogen-rich gas is extracted from the positive electrode tank of the main stack via a bypass pipeline. Electrolyte (usually positive electrode) (If the concentration exceeds 20% of the normal value), it is pumped into the cathode chamber of the rebalancing unit;

[0068] The rebalancing unit has a three-chamber, two-membrane fuel cell structure: the anode chamber uses platinum-carbon (Pt / C) with a loading of 0.5 mg / cm³. 2 The carbon paper electrode undergoes a hydrogen oxidation reaction after hydrogen gas is introduced. The cathode chamber uses graphite felt electrodes, and is circulated with... A reduction reaction occurs after electrolyte reaction ( The intermediate chamber is a sulfuric acid solution circulation chamber isolated by a proton exchange membrane (Nafion 212), used to transport protons and prevent direct mixing of hydrogen and electrolyte. During operation, the current density of the rebalancing unit is controlled at 20–60 mA / cm², running continuously for 15–30 minutes until the positive electrode electrolyte... The concentration dropped to the normal range (with) The concentration ratio is close to 1:1. The treated electrolyte is returned to the positive electrode storage tank of the main fuel cell stack via another pipeline.

[0069] During the execution of Agreement B, the main fuel cell stack negative electrode remains open-circuited, but the negative electrode electrolyte circulation pump continues to operate to continuously deliver newly generated hydrogen to the rebalancing unit. After the repair is completed, the controller closes the gas and liquid circuit valves, reconnecting the main fuel cell stack to the system.

[0070] S4.3: Final recovery verification. After executing the forced repair protocol (Protocol A and / or Protocol B), the system returns to S1, re-acquires data for at least 5 complete charge-discharge cycles, and calculates the repaired capacity decay rate according to the method in S2. and the latest cycle discharge capacity ;

[0071] If the following two conditions are met: (1) (First threshold); (2) If the system fails to recover, the third-level recovery is considered successful. The central controller switches the system operating mode back to "optimized mode" (operating parameters in S2), clears the recovery failure flags for each level, and stores all process data of this third-level recovery (including impedance spectrum, pulse parameters, and rebalancing unit operation records) in permanent memory for subsequent lifetime prediction model training.

[0072] If the above conditions are not met, the capacity degradation is determined to be irreversible. The central controller issues a "maintenance alarm" signal through the human-machine interface (HMI), prompting the operator to perform manual intervention (such as replacing the electrolyte or electrode materials), and at the same time, puts the battery system into a safe standby state to avoid further performance degradation caused by continued operation.

[0073] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A method for restoring the activity of an electrolyte in a flow battery, characterized in that, The method includes: Collect multimodal data streams of flow batteries during charge and discharge operation. The multimodal data streams include at least electrochemical data streams, physical characteristic data streams, and capacity decay characteristic data streams, and synchronize and align the multimodal data streams according to a unified timestamp. The electrolyte activity state is evaluated based on the multimodal data stream, and the capacity decay rate is calculated. When the capacity decay rate exceeds a first threshold, the first-level operating condition recovery is executed, and the operating parameters are forcibly adjusted to suppress operating condition-induced activity decay. If the capacity decay rate still exceeds the first threshold after the first-level operating condition recovery is performed, the physicochemical properties of the electrolyte are tested, and composite additives are injected according to the test results to perform the second-level chemical addition recovery to compensate for the electrolyte's intrinsic activity. If the capacity decay rate still exceeds the first threshold after performing the second-level chemical addition recovery, the electrochemical impedance spectroscopy is measured to extract the charge transfer resistance. When the charge transfer resistance exceeds a preset multiple of the initial charge transfer resistance, the third-level forced electrochemical repair is performed to restore the activity of the electrode / electrolyte interface.

2. The method for restoring electrolyte activity in a flow battery according to claim 1, characterized in that, The multimodal data stream also includes a side reaction feature data stream, which includes at least the cumulative hydrogen evolution rate. When the cumulative hydrogen evolution rate continuously exceeds the hydrogen evolution threshold, a hydrogen evolution anomaly flag is generated to trigger the short-time overpotential repair protocol in the third-level forced electrochemical repair.

3. The method for restoring electrolyte activity in a flow battery according to claim 1, characterized in that, The calculation of the capacity decay rate includes: extracting the capacity retention rate sequence of the most recent N complete charge-discharge cycles from the synchronized multimodal data stream, performing linear fitting on the capacity retention rate sequence using the least squares method, and taking the slope obtained from the fitting as the current capacity decay rate; the first threshold value ranges from 0.02% / cycle to 0.08% / cycle.

4. The method for restoring electrolyte activity in a flow battery according to claim 1, characterized in that, The first-level operating condition recovery includes: forcibly clamping the charge / discharge cutoff SOC between 0.4 and 0.6, and during the charging process, when the SOC reaches 0.6, switching the current density from a first current density to a second current density, wherein the first current density ranges from 180 to 220 mA / cm². 2 The second current density ranges from 140 to 180 mA / cm². 2 .

5. The method for restoring electrolyte activity in a flow battery according to claim 1, characterized in that, The physicochemical performance testing of the electrolyte includes: detecting the valence concentration of vanadium ions or iron / chromium ions in the electrolyte by ultraviolet-visible spectrophotometry, and / or detecting the relative proportion of active complexes in the chromium-based flow battery by Raman spectroscopy; when the relative proportion of the active complexes is less than 60%, the electrolyte is determined to be aged.

6. The method for restoring electrolyte activity in a flow battery according to claim 5, characterized in that, The composite additive includes inorganic and organic additives. The inorganic additive is selected from ferric ammonium sulfate or antimony trichloride, and the organic additive is selected from glycerol. For vanadium redox flow batteries, ferric ammonium sulfate is injected to achieve a final concentration of 0.3%–0.5% in the electrolyte. For iron-chromium flow batteries, glycerol is injected to achieve a final concentration of 0.5%–0.8%, while antimony trichloride is injected simultaneously. The final ion concentration reached 5–15 mmol / L.

7. The method for restoring electrolyte activity in a flow battery according to claim 1, characterized in that, The electrochemical impedance spectroscopy measurement involves applying a sinusoidal AC excitation signal with an amplitude of 5 mV and a frequency range of 100 kHz to 0.01 Hz to the fuel cell stack, using an equivalent circuit model. The fitting process is performed to extract the ohmic resistance, charge transfer resistance, and constant phase angle element index; the preset multiple is 2.0 times.

8. The method for restoring electrolyte activity in a flow battery according to claim 1, characterized in that, The third-level forced electrochemical repair includes an asymmetric pulse repair protocol: applying a sequence of positive and negative asymmetric high-frequency pulse voltages to the stack, with a pulse frequency of 100–500 Hz, a positive pulse amplitude of 1.2 times the rated charging voltage, a negative pulse amplitude of 0.8 times the rated charging voltage, a positive pulse width to negative pulse width ratio of 1:5, and a total repair time of 30–60 seconds, until the charge transfer resistance drops to less than 1.5 times the initial charge transfer resistance.

9. The method for restoring electrolyte activity in a flow battery according to claim 1, characterized in that, The third-level forced electrochemical repair includes a short-time overpotential repair protocol: when the battery system is an iron-chromium flow battery and the cumulative hydrogen evolution rate continuously exceeds 0.1 mL / min, the hydrogen gas evolved at the top of the negative electrode tank is introduced into the anode chamber of the hydrogen-iron rebalancing unit, enriching the positive electrode tank with hydrogen gas that has been generated due to hydrogen evolution. The electrolyte is introduced into the cathode chamber of the rebalancing unit and run for 15–30 minutes until the positive electrode electrolyte is absorbed. The concentration has dropped to the normal range.

10. The method for restoring electrolyte activity in a flow battery according to claim 1, characterized in that, The method also includes final recovery verification: after performing the third-level forced electrochemical repair, the capacity decay rate after repair and the discharge capacity of the latest cycle are recalculated. If the capacity decay rate after repair does not exceed the first threshold and the discharge capacity of the latest cycle is not less than 0.85 times the rated capacity, the recovery is determined to be successful and the system is switched back to the optimization mode; otherwise, a maintenance alarm signal is issued and the battery system is placed in a safe standby state.

Images