A kind of waste roadway compressed air energy storage leak detection device

Abstract

The application discloses a kind of abandoned roadway compressed air energy storage leak detection devices, belong to sealing detection technical field.Therein, distributed differential pressure sensor array is arranged along roadway longitudinal and ring, injection and production cycle synchronization module real-time acquisition phase information and output synchronous trigger signal, time-sharing adaptive data acquisition unit is different sampling rate according to phase window and collects differential pressure data, data analysis and early warning host extracts first response time and lag response time from the data collected, and according to the numerical combination of both, each interface damage grade is distinguished, the state of three layers of sensors is determined after logical fusion, and the degree of leakage is output.The saturated differential pressure extraction module uses the differential pressure stable value during the pressure maintaining stage to determine whether abnormal sliding occurs in the slip layer.The leakage prediction module predicts the remaining cycle number by exponential fitting and triggers the graded warning signal, realizing online monitoring and early warning of the sealing of the full interface of the composite sealing structure under non-stop state.

Description

Technical Field

[0001] This invention relates to the field of sealing detection technology, specifically a device for testing the sealing performance of compressed air energy storage in abandoned tunnels. Background Technology

[0002] Compressed air energy storage (CAES) is a large-scale energy storage technology that utilizes off-peak electricity to compress air and store it in underground space, releasing the high-pressure air to drive generators during peak electricity demand. Abandoned coal mine tunnels, due to their large underground space, can be converted into CAES storage facilities.

[0003] When abandoned coal mine roadways are converted into compressed air energy storage facilities, their sealing structures typically employ multi-layered composite structures, including steel linings, slip layers, and reinforced concrete inner linings. During periodic injection and extraction operations, the storage facility experiences alternating internal pressures, and the interfaces between structural layers may debond, slip, or crack, leading to leakage of high-pressure gas along these interfaces. Existing sealing detection methods mainly rely on shutdown pressure maintenance or passive monitoring of temperature surges during operation. The former requires interrupting normal power plant operation, while the latter only responds when a sufficient measurable signal is generated after a leak occurs, failing to provide early warning during damage development and making it difficult to pinpoint the specific interface location of the leak. How to implement continuous and effective sealing monitoring of the interfaces of the composite sealing structure without affecting normal power generation is a pressing technical problem in this field.

[0004] In view of this, the present invention proposes a device for testing the sealing performance of compressed air energy storage in abandoned tunnels, which solves the above-mentioned technical problems. Summary of the Invention

[0005] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0006] A device for testing the sealing performance of compressed air energy storage in abandoned tunnels includes: A distributed differential pressure sensor array is arranged at intervals along the longitudinal and circumferential directions of the roadway on different interfaces of the composite sealing structure of the roadway. The composite sealing structure consists of a steel liner, a slip layer, a reinforced concrete liner, and surrounding rock from the inside out. The distributed differential pressure sensor array includes a first type of differential pressure sensor embedded at the bonding interface between the steel liner and the slip layer, a second type of differential pressure sensor embedded at the contact surface between the slip layer and the reinforced concrete liner, and a third type of differential pressure sensor embedded at the contact surface between the reinforced concrete liner and the surrounding rock. The injection-production cycle synchronization module is connected to the power plant control system signal to acquire the phase information of the injection-production cycle in real time, including the start time of gas injection, the end time of gas injection, the start time of gas release, and the end time of gas release, and outputs a trigger signal synchronized with the injection-production cycle. The time-segmented adaptive data acquisition unit is signal-connected to the distributed differential pressure sensor array and the injection-import cycle synchronization module. It is used to acquire differential pressure sensor data at different sampling rates and different acquisition strategies in different phase windows according to the phase information of the injection-import cycle. The data analysis and early warning host is signal-connected to the time-segmented adaptive data acquisition unit. It is used to extract the advance response time Δt1 of each interface differential pressure sensor in the gas injection stage and the lag response time Δt2 in the gas release stage from the acquired data, and to determine the damage state of the sealing interface based on Δt1 and Δt2.

[0007] Preferably, the injection-production cycle synchronization module includes a cycle phase detection unit and a phase counter. The cycle phase detection unit is connected to at least one of the compressor start / stop signal, intake valve opening signal, and pipeline pressure signal of the power plant control system, and is used to accurately calculate the four phase nodes of the injection-production cycle. The phase counter is used to record the cumulative number of injection-production cycles n since the power plant was first put into operation, and to associate and store the number of cycles with the differential pressure detection data of each measuring point.

[0008] Preferably, the time-segmented adaptive data acquisition unit includes a pressure difference main acquisition channel, which continuously acquires differential pressure data at a sampling rate of 1kHz and stores it in a loop buffer in real time. The data is only transferred from the loop buffer to long-term storage when the differential pressure change rate exceeds a set threshold of 0.1kPa / s. In the steady-state stage when the differential pressure change rate is below the threshold, data points are stored at a low frequency.

[0009] Preferably, the data analysis and early warning host includes a lead / lag time extraction module and an interface state discrimination module. The lead / lag time extraction module extracts the lead response time Δt1 of the differential pressure at each interface during the injection stage and the lag response time Δt2 of the venting stage from the differential pressure sensor data, wherein: The advance response time Δt1 is calculated according to the following formula: Δt1=t2-t1, where t1 is the moment when the pressure at the roadway entrance begins to rise, and t2 is the moment when the pressure difference at the current measuring point first exceeds 1.05 times the benchmark pressure difference; The hysteresis response time Δt2 is calculated according to the following formula: Δt2=t4-t3, where t3 is the moment when the pressure at the roadway entrance begins to decrease, and t4 is the moment when the differential pressure value at the current measuring point begins to decrease; The interface state discrimination module uses a decision tree classifier to classify the interface state into four levels: intact, slightly damaged, moderately damaged, and severely damaged, based on different combinations of Δt1 and Δt2.

[0010] Preferably, the interface state discrimination module is further configured to: logically fuse the state discrimination results of the first type of differential pressure sensor, the second type of differential pressure sensor, and the third type of differential pressure sensor at the same measuring point; if the first type of sensor determines that the damage is moderate while the second and third types of sensors determine that the damage is intact, then the leakage is determined to occur between the steel lining and the slip layer and has not penetrated to the outside; if both the first and second types of sensors determine that the damage is moderate while the third type of sensor determines that the damage is intact, then the leakage is determined to have penetrated from the interface between the steel lining and the slip layer to the interface between the slip layer and the reinforced concrete inner lining, but has not yet penetrated to the surrounding rock interface; if the first, second, and third types of sensors simultaneously determine that the damage is moderate or above, then the leakage channel is determined to have completely penetrated all interfaces.

[0011] Preferably, the data analysis and early warning host further includes a leakage prediction module, which is configured to: read the cumulative number of cycles n recorded by the phase counter and the interface state classification history of each measuring point in the most recent M cycles; perform exponential fitting on the change of the state level of each measuring point over time: y(n)=a·e^(b·n), where y(n) is the state level, n is the normalized number of cycles, and a and b are fitting parameters; predict the remaining number of cycles Nᵣ required for the state level of the measuring point to rise from the current level to the severe damage level based on the fitting function; and issue an active early warning signal when Nᵣ is lower than a preset threshold.

[0012] Preferably, the leakage prediction module is configured to: issue a yellow warning when Nᵣ is lower than a preset first threshold, indicating that maintenance needs to be arranged soon; and issue a red warning when Nᵣ is lower than a preset second threshold, indicating that the machine should be shut down immediately and repairs should be organized.

[0013] Preferably, the data analysis and early warning host further includes a saturated differential pressure extraction module, used to extract the saturated differential pressure P of the third type of differential pressure sensor from the differential pressure stability value during the pressure holding phase. s When P s When the slip layer exceeds the design reference value, it is determined that the slip layer has experienced abnormal slippage.

[0014] Preferably, the spacing between adjacent sensing sections of the distributed differential pressure sensor array along the longitudinal direction of the roadway is 15m to 25m, and each section is equipped with sensor groups at 90° circumferential intervals, wherein the sensor density in the top area of ​​the roadway is more than twice that in the bottom area.

[0015] Preferably, the first type of differential pressure sensor has a range of -200 kPa to +200 kPa and a resolution of 0.05 Pa, the second type of differential pressure sensor has a range of -300 kPa to +300 kPa and a resolution of 0.08 Pa, and the third type of differential pressure sensor has a range of -500 kPa to +500 kPa and a resolution of 0.1 Pa; each differential pressure sensor integrates a PT100 platinum resistance temperature compensation element for temperature drift correction of the pressure measurement value.

[0016] The beneficial effects of this invention are: This invention utilizes the inherent injection-production cycle of a compressed air energy storage system as the detection excitation source, eliminating the need for additional pressure generating equipment and shutdown operations, thus achieving continuous monitoring during operation. By extracting the lead response time of the differential pressure sensor during the injection phase and the lag response time during the venting phase, it can identify sealing performance degradation in the early stages of interfacial microcrack propagation or debonding development, providing early warning of leakage risks. Simultaneously, three layers of differential pressure sensors distributed between the steel liner and the slip layer, between the slip layer and the reinforced concrete liner, and between the reinforced concrete liner and the surrounding rock can accurately locate the interface where leakage occurs and its longitudinal penetration degree. Furthermore, the saturation differential pressure analysis of the third type of differential pressure sensor during the pressure holding phase provides a quantitative basis for determining the slip layer's sliding state. The exponential fitting prediction module estimates the remaining number of cycles based on historical data of the state level, transforming maintenance decisions from reactive remediation to proactive prevention. Attached Figure Description

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

[0018] in: Figure 1 This is a schematic diagram of the overall structure of the steel liner, slip layer, reinforced concrete inner liner and surrounding rock. Figure 2 This is a schematic diagram of the cross-sectional connection structure of an abandoned alleyway; Figure 3 for Figure 2 Enlarged view of point A in the middle; Figure 4 This is a schematic diagram of the overall structure of the third type of differential pressure sensor and the surrounding rock. Figure 5 Flowchart of the injection-production cycle synchronization module; Figure 6 Flowchart of the time-segmented adaptive data acquisition unit; Figure 7Flowchart for the module of leading / lagging time extraction and interface state determination; Figure 8 Flowchart of the logic fusion and discrimination process for three-layer sensors; Figure 9 Here is a flowchart of the saturation pressure difference extraction module and the leakage prediction module; Figure 10 This is a flowchart of the overall module architecture.

[0019] In the picture: 10. Steel lining; 20. Slip layer; 30. Reinforced concrete inner lining; 40. Surrounding rock; 100. Distributed differential pressure sensor array; 110. Type I differential pressure sensor; 120. Type II differential pressure sensor; 130. Type III differential pressure sensor. Detailed Implementation

[0020] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0021] Example 1: like Figure 1 , Figure 2 , Figure 3 and Figure 10 As shown in the figure, the compressed air energy storage sealing test device for abandoned roadways described in this embodiment is applied to a compressed air energy storage tank converted from an abandoned coal mine roadway. The composite sealing structure of the tank consists of a steel lining layer 10, a sliding layer 20, a reinforced concrete inner lining layer 30, and surrounding rock 40, from the inside out.

[0022] The steel liner 10 is in direct contact with the high-pressure compressed air inside the gas storage tank, undertaking the main airtight sealing function. A sliding layer 20 is installed on the outer side of the steel liner 10. The sliding layer 20 is made of a flexible material with a low coefficient of friction. Its function is to provide relative sliding capability between the steel liner 10 and the reinforced concrete inner liner 30, releasing the deformation difference between the layers caused by temperature changes and air pressure circulation. The outermost layer is the reinforced concrete inner liner 30, which is the main load-bearing structural layer, transferring the internal pressure of the gas storage tank to the surrounding rock 40. The outermost layer is the surrounding rock 40, i.e., the natural rock mass surrounding the tunnel after excavation.

[0023] The distributed differential pressure sensor array 100 is arranged with sensing sections at intervals of 15m to 25m along the longitudinal direction of the roadway. Each section is equipped with sensor groups at 90° intervals in the circumferential direction, namely one group each at the top, bottom, left side, and right side of the roadway. Each sensor group includes a first-type differential pressure sensor 110, a second-type differential pressure sensor 120, and a third-type differential pressure sensor 130.

[0024] The first type of differential pressure sensor 110 is embedded in the bonding interface between the steel liner 10 and the slip layer 20. Under normal sealing conditions, the pore pressure at this interface is dynamically balanced with the internal pressure of the roadway, and the differential pressure reading is zero. When debonding or microcracks occur at this interface, high-pressure gas seeps into the interface gap, causing the interface pore pressure to rise, and the differential pressure sensor outputs a non-zero signal.

[0025] The second type of differential pressure sensor 120 is embedded on the contact surface between the slip layer 20 and the reinforced concrete lining layer 30 to monitor changes in pore pressure at this interface. When the slip layer 20 experiences abnormal sliding or voids appear at the interface, the differential pressure signal becomes abnormal.

[0026] The third type of differential pressure sensor 130 is embedded on the contact surface between the reinforced concrete lining 30 and the surrounding rock 40 to monitor changes in pore pressure at this interface. When a lining crack extends to the interface of the surrounding rock 40 or a crack in the loose zone of the surrounding rock 40 connects with the outside of the lining, an abnormal pressure difference occurs at this interface.

[0027] The injection-production cycle synchronization module is connected to the power plant control system signal, acquiring compressor start / stop signals, intake valve opening signals, and pipeline pressure signals in real time. It calculates the phase nodes of the injection-production cycle and outputs trigger signals synchronized with the cycle. The time-segmented adaptive data acquisition unit acquires differential pressure sensor data at different sampling rates within different phase windows based on the phase information of the injection-production cycle. The data analysis and early warning host extracts the lead and lag response times of the differential pressure sensors at each interface from the acquired data and uses this information to determine the damage status of the sealing interface.

[0028] This embodiment achieves online monitoring of the entire interface of the composite sealing structure by simultaneously deploying differential pressure sensors at three key interfaces: between the steel liner 10 and the slip layer 20, between the slip layer 20 and the reinforced concrete inner liner 30, and between the reinforced concrete inner liner 30 and the surrounding rock 40. Utilizing the injection-production cycle as a natural excitation source, no additional pressure generating equipment is required, enabling online monitoring without shutting down the system.

[0029] Example 2: like Figure 5 As shown in this embodiment, the injection-mapping cycle synchronization module includes a cycle phase detection unit and a phase counter.

[0030] The periodic phase detection unit is connected to three signal sources of the power plant control system via signal lines. The compressor start / stop signal reflects whether the compressor is running and is a switching signal. The intake valve opening signal reflects the degree of opening of the intake valve and is an analog signal ranging from 0 to 100%. The pipeline pressure signal reflects the real-time pressure value of the intake pipeline and is an analog signal ranging from 0 to 10 MPa. After preprocessing the three input signals, the periodic phase detection unit calculates the four phase nodes of the injection-production cycle.

[0031] The judgment logic for the phase node is as follows: When the compressor start signal changes from off to on, the intake valve opening signal increases from 0, and the pipeline pressure signal begins to rise, the injection start time is determined. When the compressor stop signal changes from on to off, or the intake valve opening signal decreases to 0, and the pipeline pressure signal reaches its peak and then stabilizes, the pressure holding start time is determined. When the vent valve opening signal is triggered and the pipeline pressure signal begins to decrease, the venting start time is determined. When the vent valve closes and the pipeline pressure signal drops to its lowest point and stabilizes, the venting end time is determined.

[0032] The phase counter is connected to the periodic phase detection unit. Whenever a complete injection-production cycle is detected, i.e., a complete cycle from the start of gas injection to the end of gas release, the phase counter automatically increments the cumulative cycle count and stores it. The phase counter associates the cumulative cycle count with the differential pressure detection data at each measuring point, so that each detection data has a corresponding cycle count tag.

[0033] Taking a compressed air energy storage power station in an abandoned tunnel as an example, a complete injection-production cycle of this power station takes approximately 8 hours, including 3 hours of injection, 2 hours of pressure holding, 2 hours of venting, and 1 hour of settling. When the power station control system issues a compressor start command, the cycle phase detection unit detects within 0.5 seconds that the compressor start / stop signal changes from disconnected to connected. Simultaneously, the inlet valve opening signal rises from 0, and the pipeline pressure signal rises from atmospheric pressure, thus determining the start time of injection. When the compressor stops running and the pipeline pressure stabilizes at the working pressure, the pressure holding start time is determined. When the venting valve opens and the pipeline pressure begins to decrease, the venting start time is determined. When the venting valve closes and the pipeline pressure drops to atmospheric pressure, the venting end time is determined, and the phase counter increments the cumulative cycle count by one.

[0034] This embodiment achieves accurate identification of the injection-mining cycle phase through real-time monitoring and comprehensive judgment of three signals from the power plant control system using a periodic phase detection unit. The phase counter associates and stores the cycle count with the detection data, providing a time coordinate system for long-term trend analysis and remaining life prediction.

[0035] Example 3: like Figure 6As shown in this embodiment, the time-segmented adaptive data acquisition unit includes a pressure differential main acquisition channel. The pressure differential main acquisition channel continuously acquires differential pressure signals from three types of differential pressure sensors at a sampling rate of 1kHz and stores them in a loop buffer in real time.

[0036] The main differential pressure acquisition channel performs differential calculations on the real-time acquired differential pressure data to obtain the differential pressure change rate. When the differential pressure change rate exceeds a set threshold of 0.1 kPa / s, the data is triggered to be transferred from the cyclic buffer to long-term storage, fully recording the entire process of the change in the preceding response signal. When the differential pressure change rate is below 0.1 kPa / s, only low-frequency data points are stored.

[0037] During the initial gas injection phase, the pressure inside the roadway rapidly rises from atmospheric pressure to the working pressure, and the differential pressure value increases rapidly from 0. The rate of change of differential pressure exceeds the threshold within a few seconds, triggering high-sampling-rate storage. During the pressure holding phase, the pressure inside the roadway stabilizes, and the differential pressure value stabilizes at a certain value. The rate of change of differential pressure is below the threshold, at which point only low-frequency data points are stored.

[0038] In a complete injection-production cycle, high-rate storage is triggered only during the initial injection phase, the initial venting phase, and other transient phases where the differential pressure change rate exceeds a threshold. For the remaining approximately 99% of the time, storage is performed at a low frequency. The amount of data stored in one cycle is less than one-thousandth of that stored continuously.

[0039] This embodiment uses an adaptive acquisition strategy triggered by differential pressure change rate to reduce data storage to less than one-thousandth of continuous sampling while ensuring complete recording of the entire process of changes in the leading and lagging response signals. This significantly reduces the capacity requirements and transmission bandwidth of data storage devices.

[0040] Example 4: like Figure 7 As shown in this embodiment, the data analysis and early warning host includes a lead / lag time extraction module and an interface status discrimination module.

[0041] The lead time is defined as the time difference between the moment the pressure at the roadway inlet begins to rise and the moment when the pressure difference at the current measuring point first exceeds 1.05 times the baseline pressure difference. The lead time is calculated using the following formula: Δt1 = t2 - t1, where t1 is the moment the pressure at the roadway inlet begins to rise, and t2 is the moment when the pressure difference at the current measuring point first exceeds 1.05 times the baseline pressure difference. For intact interfaces, where the interface is tightly sealed without gaps, the pressure wave is transmitted directly through the solid medium, and the lead time approaches zero. For interfaces with micro-damage, where there are tiny gaps, the pressure wave needs to compress the gas in the interface gaps before pressure transmission can begin, increasing the lead time to the order of seconds. For interfaces with penetrating fractures, the lead time further increases to the order of tens of seconds or even minutes.

[0042] The hysteresis response time is defined as the time difference between the moment the pressure at the roadway inlet begins to decrease and the moment the differential pressure at the current measuring point begins to decrease. The hysteresis response time is calculated using the following formula: Δt2 = t4 - t3, where t3 is the moment the pressure at the roadway inlet begins to decrease, and t4 is the moment the differential pressure at the current measuring point begins to decrease. For severely damaged interfaces, the gas in the interface gaps cannot be released quickly after venting, resulting in an abnormally prolonged hysteresis response time.

[0043] The interface state determination module classifies the interface state into four levels based on the numerical range of the lead time and lag time. When both the lead time and lag time are less than 50ms, the interface is considered intact, indicating a tight seal without gaps. When the lead time is between 50ms and 1s and the lag time is less than 100ms, the interface is considered slightly damaged, indicating the presence of small elastic gaps that can recover under low pressure. When the lead time is between 1s and 30s and the lag time is between 100ms and 10s, the interface is considered moderately damaged, indicating significant gaps and the beginning of gas seepage. When the lead time is greater than 30s and the lag time is greater than 10s, the interface is considered severely damaged, indicating that the gaps cannot close and large-scale gas seepage occurs.

[0044] Taking the detection data of a type I differential pressure sensor 110 at a certain measuring point during a certain injection-production cycle as an example. During the gas injection phase, the time t1 at which the roadway inlet pressure begins to rise is 08:00:00, and the time t2 at which the current measuring point's differential pressure value first exceeds 1.05 times the baseline differential pressure value is 08:00:02.5, with an advance response time Δt1 = t2 - t1 = 2.5 s. During the gas release phase, the time t3 at which the roadway inlet pressure begins to fall is 16:00:00, and the time t4 at which the current measuring point's differential pressure value begins to fall is 16:00:03.2, with a lag response time Δt2 = t4 - t3 = 3.2 s. The interface state at this measuring point is determined to be moderately damaged.

[0045] This embodiment achieves the identification of elastic opening and micro-damage signals at the sealing interface before a measurable mass leak of gas occurs by extracting the leading response time and lagging response time.

[0046] Example 5: like Figure 8 As shown, in this embodiment, the interface state discrimination module also performs logical fusion based on the state discrimination results of the first type of differential pressure sensor 110, the second type of differential pressure sensor 120 and the third type of differential pressure sensor 130 at the same measuring point to determine the longitudinal penetration degree of the leakage channel.

[0047] When the first type of differential pressure sensor 110 determines that the damage is moderate, while the second type of differential pressure sensor 120 and the third type of differential pressure sensor 130 determine that the leak is located inside the interface between the steel liner 10 and the slip layer 20, and has not yet penetrated to the outside of the slip layer 20. In this case, the repair only needs to address the debonding or micro-cracks between the steel liner 10 and the slip layer 20, without damaging the outer reinforced concrete liner 30.

[0048] When both the first type of differential pressure sensor 110 and the second type of differential pressure sensor 120 are determined to be moderately damaged, while the third type of differential pressure sensor 130 is determined to be intact, it is determined that the leak has penetrated from the interface between the steel liner 10 and the slip layer 20 to the interface between the slip layer 20 and the reinforced concrete inner liner 30, but has not yet penetrated to the interface of the surrounding rock 40. Repair requires simultaneous treatment of both interfaces.

[0049] When the first type of differential pressure sensor 110, the second type of differential pressure sensor 120 and the third type of differential pressure sensor 130 are all determined to be moderately damaged or above, it is determined that the leakage channel has completely penetrated all three interfaces, that is, it runs directly from the inside of the steel liner 10 to the surrounding rock 40, and the machine needs to be stopped immediately for comprehensive repair.

[0050] Taking the detection data of a certain sensing section as an example, the first type of differential pressure sensor 110 has an advance response time of 2.5s and a lag response time of 3.2s, and is judged to be moderately damaged. The second type of differential pressure sensor 120 has an advance response time of 0.08s and a lag response time of 0.06s, and is judged to be intact. The third type of differential pressure sensor 130 has an advance response time of 0.05s and a lag response time of 0.04s, and is judged to be intact. According to the logic fusion discrimination rules, the interface state discrimination module determines that the leakage occurs inside the interface between the steel liner 10 and the slip layer 20, and has not yet penetrated to the outside. Based on this, maintenance personnel can formulate a precise repair plan, which only requires local repair of the interface between the steel liner 10 and the slip layer 20, without damaging the reinforced concrete inner liner 30.

[0051] This embodiment achieves precise positioning of the longitudinal continuity of the leakage channel by logically fusing the status judgment results of three layers of sensors, providing a precise basis for the formulation of repair plans.

[0052] Example 6: like Figure 9As shown, in this embodiment, the data analysis and early warning host also includes a leakage prediction module. The leakage prediction module reads the cumulative number of cycles recorded by the phase counter and the interface state classification history of each measuring point in the most recent cycles, and performs exponential fitting on the change of the state level of each measuring point over time. The form of the fitting function is that the state level y is equal to the fitting parameter a multiplied by the natural constant e raised to the power of (fitting parameter b multiplied by the normalized number of cycles n), i.e., y = a·e^(b·n), where the state level takes the values ​​0, 1, 2, and 3, and the normalized number of cycles is the cumulative number of cycles divided by the total number of designed cycles.

[0053] Let's take historical detection data from a certain measuring point as an example. The measuring point's status level is Level 1 after 1000 cycles, Level 1 after 2000 cycles, Level 2 after 3000 cycles, Level 2 after 4000 cycles, and Level 2 after 5000 cycles. The leakage prediction module performs exponential fitting on these data points, obtaining fitting parameters a=0.98 and b=2.1. The current cycle count for this measuring point is 5000, and the designed total cycle count is 10000. Therefore, the current normalized cycle count is 5000 / 10000=0.5. Substituting the fitting parameters into the fitting function, and setting y=3, the calculation is as follows: 3 = 0.98 × e^(2.1 × n); e^(2.1 × n) = 3 / 0.98 ≈ 3.061; 2.1 × n = ln(3.061) ≈ 1.119; n ≈ 0.533.

[0054] When the damage level reaches Level 3 (severe damage), the corresponding normalized cycle count is approximately 0.533. Subtracting the predicted normalized cycle count from the current normalized cycle count yields a difference of approximately 0.033. Multiplying this difference by the designed total number of cycles (10,000) gives a predicted remaining cycle count of approximately 330. This means the measurement point is predicted to evolve from moderate damage to severe damage after approximately 330 cycles.

[0055] This embodiment achieves quantitative prediction of the remaining life of the sealing interface by exponentially fitting historical data of the status level, thus transforming the operation and maintenance mode from post-repair to pre-prevention.

[0056] Example 7: like Figure 9 As shown, in this embodiment, the leakage prediction module triggers different levels of early warning signals based on the comparison result of the predicted remaining number of cycles with a preset threshold.

[0057] A yellow alert is triggered when the predicted remaining number of cycles is less than 100, indicating that the sealing interface damage at the measuring point has progressed to the late stage of moderate damage and is expected to evolve into severe damage within 100 cycles. Maintenance personnel should arrange for a detailed inspection and preventive maintenance of the area within one to two weeks.

[0058] A red alert is triggered when the predicted remaining number of cycles is less than 20, indicating that the sealing interface damage at the measuring point is approaching the critical point of severe damage, and gas leakage is expected to occur within 20 cycles. The operation and maintenance personnel should immediately stop the operation of the gas storage facility and organize repairs.

[0059] Following the calculation example in Example 6, if the predicted remaining cycles for a certain measuring point are 330, which is greater than 100, no warning will be triggered. If the predicted remaining cycles for another measuring point are 75, which is less than 100 but greater than 20, a yellow warning will be triggered. The ground monitoring terminal will display the measuring point's location marked in yellow and suggest arranging for immediate maintenance. If the predicted remaining cycles for a third measuring point are 15, which is less than 20, a red warning will be triggered. The ground monitoring terminal will display the measuring point's location marked in red and suggest immediate shutdown for repair, while simultaneously triggering an audible and visual alarm.

[0060] This embodiment uses a two-level early warning mechanism to transform the quantitative remaining life prediction results into clear operation and maintenance instructions, enabling operation and maintenance personnel to allocate resources reasonably according to the degree of urgency.

[0061] Example 8: like Figure 4 , Figure 9 As shown, in this embodiment, the data analysis and early warning host also includes a saturated differential pressure extraction module. The saturated differential pressure extraction module extracts the saturated differential pressure of the third type of differential pressure sensor 130 from the stable differential pressure value during the pressure holding stage.

[0062] During the pressure holding stage, the pressure inside the roadway is maintained at the rated working pressure, and the pore pressure at each interface gradually stabilizes. The third type of differential pressure sensor 130 measures the difference between the pore pressure at the interface between the reinforced concrete lining layer 30 and the surrounding rock 40 and the pressure inside the roadway. This difference, after stabilizing during the pressure holding stage, becomes the saturation pressure difference P. s .

[0063] The saturation pressure difference reflects the ability of gas to penetrate each layer and reach the interface of the surrounding rock 40 under stable high pressure. The higher the saturation pressure difference, the worse the sealing function of the slip layer 20. When the saturation pressure difference exceeds the design reference value of the slip layer 20, it is determined that the slip layer 20 has experienced abnormal sliding. The abnormal sliding of the slip layer 20 is manifested by excessive shear deformation of the slip layer 20 material, forming gas seepage channels, which causes an abnormal increase in pore pressure at the interface between the reinforced concrete lining layer 30 and the surrounding rock 40.

[0064] The design baseline value of the slip layer 20 in an abandoned tunnel compressed air energy storage power station is 20 kPa. During the first pressure holding phase, the differential pressure value of the third-type differential pressure sensor 130 stabilizes at 35 kPa approximately 30 minutes after the pressure holding begins. The saturated differential pressure extraction module compares 35 kPa with the design baseline value of 20 kPa. Since 35 kPa is greater than 20 kPa, exceeding the design baseline value by 75%, it determines that the slip layer 20 has experienced abnormal slippage and outputs the result to the ground monitoring terminal. After receiving the abnormal slippage alarm for the slip layer 20, maintenance personnel can specifically check the slippage status of the slip layer 20 and replace or reinforce it if necessary.

[0065] This embodiment achieves quantitative detection of the sliding state of the slip layer 20 by monitoring the saturated differential pressure during the pressure holding stage using a third type of differential pressure sensor 130.

[0066] Example 9: like Figure 1 , Figure 2 and Figure 4 As shown, in this embodiment, the spacing between adjacent sensing sections of the distributed differential pressure sensor array 100 along the longitudinal direction of the roadway is 15m to 25m, and each section is equipped with sensor groups at 90° intervals around the circumference, namely one group each at the top, bottom, left waist, and right waist.

[0067] The density of sensors deployed at the top of the tunnel is more than twice that at the bottom. Specifically, two sets of sensors are deployed within a 90° range at the top, namely slightly to the left and slightly to the right, while only one set of sensors is deployed within a 90° range at the bottom, namely at the center.

[0068] The basis for this asymmetric deployment strategy is that the stability of the top surrounding rock 40 in abandoned tunnels is weaker than that of the side and bottom surrounding rock 40 after excavation. Tunnel excavation leads to stress redistribution in the surrounding rock 40, with the top surrounding rock 40 being in a tensile stress zone, making it more prone to tensile cracks and loosening rings. Therefore, the top area is a high-risk area for sealing failure and requires higher detection resolution.

[0069] Taking a 200m long abandoned tunnel as an example, the longitudinal sensing cross-sections are spaced 20m apart, with a total of 11 cross-sections. Each cross-section has two sets of sensors at the top, one set at the bottom, one set on the left side, and one set on the right side. Each sensor set contains three differential pressure sensors. The total number of sensors in the top area is 66, and the number of sensors in the bottom area is 33. The density at the top is twice that at the bottom.

[0070] This embodiment achieves higher detection resolution and prediction accuracy in the weakest area by densely deploying sensors in the top region, while controlling the total number of sensors and system cost.

[0071] Example 10: like Figure 1 , Figure 2 and Figure 3 As shown, in this embodiment, the three types of differential pressure sensors are configured differently according to the permeability characteristics of their respective interfaces. The first type of differential pressure sensor 110 has a range of -200 kPa to +200 kPa and a resolution of 0.05 Pa, because the permeability difference between the steel liner 10 and the slip layer 20 is the smallest, and the pressure difference change is the weakest, requiring the highest accuracy. The second type of differential pressure sensor 120 has a range of -300 kPa to +300 kPa and a resolution of 0.08 Pa. The third type of differential pressure sensor 130 has a range of -500 kPa to +500 kPa and a resolution of 0.1 Pa, because the permeability difference between the reinforced concrete liner 30 and the surrounding rock 40 is the largest, and the pressure change is the greatest.

[0072] Each differential pressure sensor integrates a PT100 platinum resistance temperature compensation element. The output signal of the differential pressure sensor is affected by temperature and will drift. The PT100 detects the temperature at the sensor's location in real time, and the signal processing circuit corrects the differential pressure measurement value based on the temperature value.

[0073] Taking the measurement data of a type 1 differential pressure sensor 110 at a certain measuring point during the gas injection stage as an example, the current temperature is 45℃, the reference temperature is 20℃, and the measured differential pressure value is 102.5 Pa. After temperature correction, the differential pressure value is 101.7 Pa, which differs from the uncorrected 102.5 Pa by 0.8 Pa. Without correction, the 0.8 Pa error will affect the accurate extraction of the lead time, especially when the differential pressure value is close to the critical point of 1.05 times the reference differential pressure.

[0074] This embodiment utilizes differentiated measurement range configurations for three types of sensors to match the detection sensitivity of each interface with its respective pressure change amplitude. Temperature compensation elements eliminate the impact of temperature fluctuations within the roadway on measurement accuracy, ensuring the accuracy of the detection data.

[0075] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of protection claimed by the present invention. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A device for testing the sealing performance of compressed air energy storage in abandoned tunnels, characterized in that, include: A distributed differential pressure sensor array (100) is arranged at intervals along the longitudinal and circumferential directions on different interfaces of the composite sealing structure of the roadway. The composite sealing structure consists of a steel liner (10), a slip layer (20), a reinforced concrete inner liner (30), and surrounding rock (40) from the inside out. The distributed differential pressure sensor array (100) includes a first type of differential pressure sensor (110) embedded at the bonding interface between the steel liner (10) and the slip layer (20), a second type of differential pressure sensor (120) embedded at the contact surface between the slip layer (20) and the reinforced concrete inner liner (30), and a third type of differential pressure sensor (130) embedded at the contact surface between the reinforced concrete inner liner (30) and the surrounding rock (40). The injection-production cycle synchronization module is connected to the power plant control system signal to acquire the phase information of the injection-production cycle in real time, including the start time of gas injection, the end time of gas injection, the start time of gas release, and the end time of gas release, and outputs a trigger signal synchronized with the injection-production cycle. The time-segmented adaptive data acquisition unit is signal-connected to the distributed differential pressure sensor array (100) and the injection-import cycle synchronization module, and is used to acquire differential pressure sensor data at different sampling rates and different acquisition strategies in different phase windows according to the phase information of the injection-import cycle. The data analysis and early warning host is signal-connected to the time-segmented adaptive data acquisition unit. It is used to extract the advance response time Δt1 of each interface differential pressure sensor in the gas injection stage and the lag response time Δt2 in the gas release stage from the acquired data, and to determine the damage state of the sealing interface based on Δt1 and Δt2.

2. The compressed air energy storage sealing test device for abandoned tunnels according to claim 1, characterized in that, The injection-production cycle synchronization module includes a cycle phase detection unit and a phase counter. The cycle phase detection unit is connected to at least one of the compressor start / stop signal, intake valve opening signal and pipeline pressure signal of the power plant control system, and is used to accurately calculate the four phase nodes of the injection-production cycle. The phase counter is used to record the cumulative number of injection-production cycles n since the power plant was first put into operation, and to associate and store the number of cycles with the differential pressure detection data of each measuring point.

3. The compressed air energy storage sealing test device for abandoned tunnels according to claim 1, characterized in that, The time-segmented adaptive data acquisition unit includes a pressure difference main acquisition channel. The pressure difference main acquisition channel continuously acquires differential pressure data at a sampling rate of 1kHz and stores it in a loop buffer in real time. Data is only transferred from the loop buffer to long-term storage when the differential pressure change rate exceeds a set threshold of 0.1kPa / s. In the steady-state stage when the differential pressure change rate is below the threshold, data points are stored at a low frequency.

4. The compressed air energy storage sealing test device for abandoned tunnels according to claim 1, characterized in that, The data analysis and early warning host includes a lead / lag time extraction module and an interface state discrimination module. The lead / lag time extraction module extracts the lead response time Δt1 of the differential pressure at each interface during the gas injection stage and the lag response time Δt2 of the gas release stage from the differential pressure sensor data, wherein: The advance response time Δt1 is calculated according to the following formula: Δt1=t2-t1, where t1 is the moment when the pressure at the roadway entrance begins to rise, and t2 is the moment when the pressure difference at the current measuring point first exceeds 1.05 times the benchmark pressure difference; The hysteresis response time Δt2 is calculated according to the following formula: Δt2=t4-t3, where t3 is the moment when the pressure at the roadway entrance begins to decrease, and t4 is the moment when the differential pressure value at the current measuring point begins to decrease; The interface state discrimination module uses a decision tree classifier to classify the interface state into four levels: intact, slightly damaged, moderately damaged, and severely damaged, based on different combinations of Δt1 and Δt2.

5. The compressed air energy storage sealing test device for abandoned tunnels according to claim 4, characterized in that, The interface state discrimination module is further configured to: perform logical fusion based on the state discrimination results of the first type of differential pressure sensor (110), the second type of differential pressure sensor (120) and the third type of differential pressure sensor (130) at the same measuring point; if the first type of sensor determines that the damage is moderate while the second and third types of sensors determine that the damage is intact, then the leakage is determined to occur between the steel lining (10) and the slip layer (20) and has not penetrated to the outside; if both the first and second types of sensors determine that the damage is moderate while the third type of sensor determines that the damage is intact, then the leakage is determined to have penetrated from the interface between the steel lining (10) and the slip layer (20) to the interface between the slip layer (20) and the reinforced concrete inner lining (30), but has not yet penetrated to the interface of the surrounding rock (40); if the first, second and third types of sensors simultaneously determine that the damage is moderate or above, then the leakage channel is determined to have completely penetrated all interfaces.

6. The compressed air energy storage sealing test device for abandoned tunnels according to claim 1, characterized in that, The data analysis and early warning host also includes a leakage prediction module. The leakage prediction module is configured to: read the cumulative number of cycles n recorded by the phase counter and the interface state classification history of each measuring point in the most recent M cycles; perform exponential fitting on the change of the state level of each measuring point over time: y(n)=a·e^(b·n), where y(n) is the state level, n is the normalized number of cycles, and a and b are fitting parameters; predict the remaining number of cycles Nᵣ required for the state level of the measuring point to rise from the current level to the severe damage level based on the fitting function; and issue an active early warning signal when Nᵣ is lower than a preset threshold.

7. The compressed air energy storage sealing test device for abandoned tunnels according to claim 6, characterized in that, The leakage prediction module is configured to issue a yellow warning when Nᵣ is lower than a preset first threshold, indicating that maintenance needs to be arranged soon; and issue a red warning when Nᵣ is lower than a preset second threshold, indicating that the machine should be shut down immediately and repairs should be organized.

8. The compressed air energy storage sealing test device for abandoned tunnels according to claim 1, characterized in that, The data analysis and early warning host also includes a saturated differential pressure extraction module, used to extract the saturated differential pressure P of the third type of differential pressure sensor (130) from the differential pressure stability value during the pressure holding phase. s When P s When the slip layer (20) exceeds the design reference value, it is determined that the slip layer (20) has experienced abnormal slippage.

9. The compressed air energy storage sealing test device for abandoned tunnels according to claim 1, characterized in that, The distributed differential pressure sensor array (100) has an adjacent sensing section spacing of 15m to 25m along the longitudinal direction of the roadway, and each section is equipped with sensor groups at 90° circumferential intervals. The sensor density in the top area of ​​the roadway is more than twice that in the bottom area.

10. The compressed air energy storage sealing test device for abandoned tunnels according to claim 1, characterized in that, The first type of differential pressure sensor (110) has a range of -200kPa to +200kPa and a resolution of 0.05Pa, the second type of differential pressure sensor (120) has a range of -300kPa to +300kPa and a resolution of 0.08Pa, and the third type of differential pressure sensor (130) has a range of -500kPa to +500kPa and a resolution of 0.1Pa. Each differential pressure sensor integrates a PT100 platinum resistance temperature compensation element for temperature drift correction of the pressure measurement value.

Images