Communication salt cavern gas storage operation scheduling method and system

Abstract

The invention relates to the technical field of gas storage operation scheduling, and discloses a communicated salt cavern gas storage operation scheduling method and system, and the method comprises the steps: receiving real-time pressure data and injection-production flow data from a plurality of salt caverns; according to the data, the current fluid transmission efficiency among the salt caverns is reversely calculated; tracking a time-varying track of the fluid transmission efficiency, identifying an evolution trend of the fluid transmission efficiency according to the track, and establishing an evolution function of the fluid transmission efficiency; simulating a future operation scene based on the current fluid transmission efficiency, the evolution function of the fluid transmission efficiency and external natural gas demand prediction, and pre-judging a potential operation risk according to a simulation result; and when the potential operation risk is pre-judged, early warning is triggered, and an injection-production plan is adjusted according to the potential operation risk. According to the invention, refined and intelligent scheduling of the operation of the communicated salt cavern gas storage is realized, and the safety, reliability and economic benefits of the operation of the gas storage are remarkably improved.

Description

Technical Field

[0001] This application relates to the field of gas storage operation and scheduling technology, and more specifically, to a method and system for operation and scheduling of interconnected salt cavern gas storage facilities. Background Technology

[0002] Interconnected salt cavern gas storage systems play a crucial role in energy infrastructure, balancing market supply and demand and ensuring energy stability through natural gas injection, extraction, and inter-storage allocation. The operation and scheduling of such systems typically rely on a fundamental model describing the working characteristics of the storage cluster, guiding injection and extraction operations to maximize economic benefits. However, during long-term operation, the complexity of the geological environment and the continuous impact of the injection and extraction cycle cause slow and subtle deformation of the surrounding rock of the salt caverns, altering the actual fluid connectivity between them. This change is implicit and dynamic, causing the initially established static scheduling model to gradually deviate from the actual physical state of the storage facilities. This can lead to serious operational risks and safety hazards, especially under extreme conditions requiring high-intensity emergency gas extraction.

[0003] Under the aforementioned scheduling framework, a key assumption is that the parameters describing the interactions between salt caverns remain essentially constant. When calculating the gas extraction rate of a salt cavern, the scheduling model considers the pressure support exerted by adjacent salt caverns, which is considered a relatively fixed and predictable quantity. However, this is not the case during the long service life of a gas storage facility, which can last for years or even decades. Each injection and extraction is equivalent to a cyclic loading and unloading of pressure and temperature on the surrounding salt rock mass. This long-term, repeated mechanical disturbance causes minute, slow creep and deformation in this special rock mass. These microscopic geomechanical changes gradually accumulate and eventually manifest in the macroscopic operating characteristics of the gas storage facility. This change is slow and implicit, and daily operational data fluctuations may not clearly reflect this trend. If the scheduling system still uses the fixed model established based on the initial state, the error between the model's predictions and the actual response of the gas storage facility will subtly increase. Such deviations may not be noticeable during periods of stable operation, but they will become apparent under extreme conditions and lead to serious consequences. Summary of the Invention

[0004] The purpose of this invention is to address the problem that during the long-term operation of interconnected salt cavern gas storage facilities, the dynamic changes in the fluid connectivity characteristics between salt caverns due to the complexity of the geological environment and the influence of injection and production cycles cause deviations between existing static scheduling models and the actual physical state, which may lead to serious operational risks and safety hazards, especially under extreme operating conditions. The invention proposes an operation scheduling method and system for interconnected salt cavern gas storage facilities.

[0005] This invention is achieved through the following technical solution: A method for operating and scheduling interconnected salt cavern gas storage facilities includes the following steps: S1. Receive real-time pressure data and injection / production flow rate data from multiple salt caverns; S2. Based on the real-time pressure data and the injection / production flow rate data, the current fluid transfer efficiency between the multiple salt caverns is calculated in reverse. S3. Track the trajectory of the fluid transport efficiency over time, and identify the evolution trend of the fluid transport efficiency based on the trajectory, thereby establishing the evolution function of the fluid transport efficiency; S4. Based on the current fluid transmission efficiency, the evolution function of the fluid transmission efficiency, and the external natural gas demand forecast, simulate future operating scenarios and predict potential operating risks based on the simulation results; S5. When the potential operational risks are anticipated, an early warning is triggered, and the injection and production plan is adjusted according to the potential operational risks.

[0006] A system for operating and scheduling interconnected salt cavern gas storage facilities includes: The data receiving module is used to receive real-time pressure data and injection / production flow data from multiple salt caverns; The efficiency calculation module is used to reverse-calculate the current fluid transfer efficiency between the multiple salt caverns based on the real-time pressure data and the injection-production flow rate data. An evolution function establishment module is used to track the trajectory of the fluid transport efficiency changing over time, and to identify the evolution trend of the fluid transport efficiency based on the trajectory, thereby establishing the evolution function of the fluid transport efficiency; The scenario simulation and risk prediction module is used to simulate future operating scenarios based on the current fluid transmission efficiency, the evolution function of the fluid transmission efficiency, and external natural gas demand forecasts, and to predict potential operating risks based on the simulation results. The plan adjustment module is used to trigger an early warning when the potential operational risks are anticipated, and to adjust the injection and production plan according to the potential operational risks.

[0007] The present invention has the following beneficial effects: By introducing dynamic calculation of fluid transmission efficiency, establishment of evolution function, and scenario simulation and risk prediction mechanism based on multiple factors, the present invention realizes refined and intelligent scheduling of interconnected salt cavern gas storage operation, significantly improves the safety, reliability and economic benefits of gas storage operation, and overcomes many problems caused by the inability of static models in the prior art to adapt to the dynamically changing interconnected characteristics of salt caverns. Attached Figure Description

[0008] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0010] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0011] To better understand the operation and scheduling method for interconnected salt cavern gas storage proposed in this application, it is necessary to explain some of the key terms involved.

[0012] "Salt caverns" are underground gas storage spaces formed by dissolving salt layers. They are characterized by good sealing and large gas storage capacity, and are important facilities for underground natural gas storage.

[0013] "Real-time pressure data" refers to the pressure values ​​continuously collected by sensors installed inside the salt cavern or in the wellbore during the operation of the salt cavern. These data reflect the instantaneous pressure status of the natural gas inside the salt cavern.

[0014] "Injection and production flow data" refers to the instantaneous flow rate of natural gas injected into or extracted from the salt cavern during its operation. These data reflect the speed and scale of natural gas entering and leaving the salt cavern.

[0015] "Fluid transport efficiency" refers to the ease with which natural gas flows between different salt caverns or between a salt cavern and an external pipeline network in a connected salt cavern gas storage system. It is affected by various factors such as geological structure, surrounding rock deformation, and fluid properties. The dynamic changes in fluid transport efficiency are one of the core issues addressed in this application.

[0016] An "evolution function" is a mathematical model that describes how fluid transport efficiency changes over time, and this function can predict the future trend of fluid transport efficiency. "External natural gas demand forecasting" refers to the prediction of natural gas consumption over a future period based on factors such as market supply and demand, seasonal variations, and meteorological conditions. This is an important basis for formulating injection and production plans.

[0017] "Operational scenario simulation" refers to simulating the operating status and performance of a salt cavern gas storage system under different injection and production plans and external demand conditions by establishing a mathematical model.

[0018] "Potential operational risks" refer to potential safety hazards, economic losses, or system failures that may occur during the operation of salt cavern gas storage facilities, such as instability of the surrounding rock of the salt cavern or interruption of gas supply.

[0019] The "injection and production plan" refers to the specific arrangements for the injection and production of natural gas in a salt cavern gas storage facility, including the injection and production volume, injection and production rate, and injection and production time.

[0020] The core of the operation and scheduling method for interconnected salt cavern gas storage proposed in this application lies in the dynamic perception, prediction, and intelligent decision-making based on the fluid transfer efficiency between salt caverns.

[0021] Please see Figure 1 As shown, a method for operation and scheduling of a connected salt cavern gas storage facility includes the following steps: S1. Receive real-time pressure data and injection / production flow rate data from multiple salt caverns; First, the method involves receiving real-time pressure and injection / production flow data from multiple salt caverns. This data is fundamental to understanding the operational status of the salt cavern system. For example, this data can be acquired in real time by installing high-precision pressure sensors and flow meters at the wellhead of each salt cavern. These sensors can be configured to automatically record data every few seconds or minutes and transmit it to a central data processing unit via wired or wireless network. In practice, this data may contain noise or outliers, so preliminary cleaning and verification are usually required after receipt. This can be done by setting reasonable threshold ranges to filter out obviously erroneous readings, or by using methods such as moving averages to smooth the data.

[0022] S2. Based on the real-time pressure data and the injection / production flow rate data, the current fluid transfer efficiency between the multiple salt caverns is calculated in reverse. Secondly, based on real-time pressure and injection / production flow data, the current fluid transport efficiency between multiple salt caverns is calculated in reverse. Fluid transport efficiency is a key indicator of the connectivity between salt caverns. One approach is to use numerical simulation software, taking real-time collected pressure and flow data as input, and iteratively calculating to find the fluid transport efficiency parameters that best fit these observation data. For example, a subsurface fluid flow simulator based on the finite element method or finite volume method can be used to abstract the salt cavern system into a network consisting of multiple nodes (salt caverns) and edges (fluid channels) connecting these nodes. By adjusting the transport efficiency parameters of each edge, the error between the simulated pressure and flow changes and the actual observed data is minimized. Another approach is to use a data-driven method, such as using a machine learning model, taking historical pressure and flow data and the corresponding fluid transport efficiency as training samples, and learning the mapping relationship from pressure and flow data to fluid transport efficiency. When new real-time data is input, the model can directly predict the current fluid transport efficiency.

[0023] S3. Track the trajectory of the fluid transport efficiency over time, and identify the evolution trend of the fluid transport efficiency based on the trajectory, thereby establishing the evolution function of the fluid transport efficiency; Secondly, the trajectory of fluid transport efficiency over time is tracked, and the evolutionary trend of fluid transport efficiency is identified based on the trajectory, thereby establishing an evolutionary function for fluid transport efficiency. Fluid transport efficiency is not static but changes slowly with factors such as operating time and geological stress. One approach is to store historically calculated fluid transport efficiency data as a time series and process this data using time series analysis methods, such as moving averages, exponential smoothing, or Kalman filtering, to identify its long-term evolutionary trend. For example, one can observe whether fluid transport efficiency exhibits a continuous upward or downward trend, or whether there are periodic fluctuations. Based on this, machine learning models such as regression analysis, neural networks, or Gaussian processes can be used to establish an evolutionary function that can describe and predict future changes in fluid transport efficiency. This function can be a simple linear function or a complex model that considers nonlinear factors.

[0024] S4. Based on the current fluid transmission efficiency, the evolution function of the fluid transmission efficiency, and the external natural gas demand forecast, simulate future operating scenarios and predict potential operating risks based on the simulation results; One approach to this step is to use the currently calculated fluid transport efficiency as the initial state of the simulation, employ an evolution function to predict the fluid transport efficiency at different future time points, and use external natural gas demand forecasts as boundary conditions for the simulation. For example, various external demand scenarios (such as high demand, medium demand, and low demand) can be set, and a dynamic gas storage system model can be used to simulate each scenario. This model considers the gas storage capacity of each salt cavern, wellbore flow characteristics, and fluid transport efficiency between salt caverns. Through simulation, key indicators such as salt cavern pressure changes, gas production capacity, and gas supply stability can be predicted under different injection and production plans. Based on the simulation results, potential operational risks that may lead to excessively low salt cavern pressure, insufficient gas production capacity, or abnormal surrounding rock stress can be identified.

[0025] S5. When the potential operational risks are anticipated, an early warning is triggered, and the injection and production plan is adjusted according to the potential operational risks.

[0026] Finally, when potential operational risks are anticipated, an early warning is triggered, and the injection and production plan is adjusted accordingly. Early warning mechanisms and plan adjustments are crucial aspects of risk management. One approach is to set a series of risk thresholds, such as minimum safe pressure in salt caverns, maximum gas production rate, and the rate of change of surrounding rock stress. When scenario simulation results indicate that any one or more indicators may exceed these thresholds, the system automatically triggers an early warning and notifies operations personnel via audible and visual alarms, SMS notifications, or emails. Simultaneously, the system automatically or semi-automatically generates a revised injection and production plan based on the specific type and severity of the anticipated risk. For example, if the pressure in a salt cavern is predicted to be too low, the system may suggest reducing the gas production rate of that cavern and increasing the gas production of adjacent caverns to compensate, or it may suggest early injection to increase the pressure in the storage area. These adjusted plans are then validated through scenario simulation again to ensure that the new plan effectively mitigates risks and meets operational objectives.

[0027] The operational scheduling method for interconnected salt cavern gas storage proposed in this application forms a dynamic and intelligent operational scheduling closed loop through the close coordination of the aforementioned technical features. First, by continuously receiving real-time pressure and injection / production flow data, the system provides the most basic and timely information input for the entire scheduling system. This data serves as the "eyes" for understanding the current operating status of the salt cavern system. Next, the efficiency calculation module uses this real-time data to inversely calculate the current fluid transfer efficiency between multiple salt caverns. This step is one of the core innovations of this application, breaking through the limitations of traditional static models and enabling the system to perceive the dynamic changes in connectivity between salt caverns in real time. Subsequently, the evolution function establishment module tracks the trajectory of fluid transfer efficiency over time based on long-term historical data and identifies the evolution trend, thereby establishing an evolution function for fluid transfer efficiency. This evolution function is key to predicting future changes in salt cavern connectivity, enabling the system to move from "seeing the present" to "foreseeing the future."

[0028] It should be further explained that, in actual operation, the evolution of fluid transport efficiency in step S3 may be affected by a variety of complex factors, such as sensor drift and changes in local geological stress. If these factors are not effectively identified and addressed, the accuracy of the evolution function may be insufficient, thereby affecting the reliability of subsequent operational scenario simulations and risk predictions. To address this, this application further proposes an optimization of the steps described above: tracking the trajectory of the fluid transport efficiency over time and identifying the evolution trend of the fluid transport efficiency based on the trajectory to establish the evolution function of the fluid transport efficiency, thereby improving the accuracy and robustness of the evolution function establishment.

[0029] Specifically, step S3 includes: S311, Receive real-time pressure data, injection-production flow rate data, and auxiliary data related to local geological stress from multiple salt caverns; In addition to conventional pressure and flow monitoring data, this invention also acquires auxiliary information reflecting changes in the geological stress state of the surrounding rock of salt caverns. This auxiliary data may include, but is not limited to, microseismic monitoring data, formation stress measurement data, rock mechanics parameter variation data, or stress field distribution data obtained through geological model simulation. This data can be obtained using dedicated geological sensors, distributed fiber optic sensing systems, or numerical simulation tools. The aim is to provide deeper geological background information on the evolution of fluid transport efficiency, thereby enabling a more accurate understanding and prediction of its changes.

[0030] S312. Based on the real-time pressure data and the injection / production flow rate data, simultaneously optimize the fluid transmission efficiency parameter and the sensor drift parameter; Specifically, a joint optimization algorithm, such as extended Kalman filtering, particle filtering, or Bayesian inference-based methods, is used to model and correct for potential drift errors in the sensor while estimating fluid transport efficiency parameters.

[0031] S313. Based on the auxiliary data, adjust the preference for changes in the fluid transport efficiency parameter and the sensor drift parameter; Specifically, during the optimization process, auxiliary data is used as prior information or constraints to dynamically adjust the model's sensitivity or confidence level to changes in fluid transport efficiency parameters and sensor drift parameters. For example, when auxiliary data (such as increased microseismic activity) indicates that significant changes may occur in the local geological structure, the model may be guided to accept larger changes in fluid transport efficiency parameters, while maintaining a relatively conservative estimate of changes in sensor drift parameters, and vice versa.

[0032] S314. Assign confidence weights to the fluid transport efficiency parameters; Specifically, based on factors such as the source of the fluid transport efficiency parameter, measurement accuracy, data quality, degree of conformity with the physical model, and consistency with auxiliary data, a weight value reflecting its reliability is assigned to each parameter.

[0033] S315. Learn the evolution function from the confidence-weighted fluid transport efficiency parameters; Specifically, the evolution function refers to the function of fluid transport efficiency constructed and trained using machine learning algorithms or time series analysis methods, such as recurrent neural networks (RNN), long short-term memory networks (LSTM), Gaussian process regression or autoregressive integral moving average models (ARIMA), with weighted fluid transport efficiency parameters as input.

[0034] S316. When a persistent deviation occurs between the trend predicted by the evolution function and the current high-confidence fluid transport efficiency parameter, a deviation-driven reassessment is triggered.

[0035] In some preferred embodiments, a connected salt cavern gas storage system is assumed, comprising three salt caverns: 100, 200, and 300. To establish a more accurate fluid transport efficiency evolution function, the system receives not only conventional real-time pressure and injection / production flow data but also microseismic monitoring data from the salt cavern area as auxiliary data related to local geological stress. When microseismic activity intensifies, indicating potential changes in local geological stress, the system adjusts its preference for changes in the fluid transport efficiency parameter to better reflect actual geological structural changes rather than simple sensor drift. Simultaneously, the fluid transport efficiency parameter and sensor drift parameter are jointly estimated using a Kalman filter algorithm, and confidence weights are assigned to the fluid transport efficiency parameter based on sensor calibration records and data consistency. For example, readings from newly calibrated sensors may be assigned a higher confidence level. Subsequently, a model based on a Long Short-Term Memory (LSTM) network is used to learn the fluid transport efficiency evolution function from these confidence-weighted parameters. If, over several consecutive weeks, there is a persistent deviation of more than 5% between the fluid transport efficiency trend predicted by the LSTM model and the actual observed values ​​of the high-confidence parameters, the system will automatically trigger a deviation-driven reassessment, which may include retraining the LSTM model or requesting intervention from human experts to ensure the accuracy of the evolution function.

[0036] If step S5 is implemented based solely on immediate risk indicators when adjusting the injection-production plan, it may fail to adequately consider the long-term geomechanical response of the salt cavern surrounding rock, such as the cumulative changes in stress state and creep rate. These long-term effects may lead to a decrease in the stability of the surrounding rock structure and even alter the long-term evolution trend of fluid transport efficiency between salt caverns, thereby introducing new and deeper operational risks. Therefore, step S5 of this application includes: S51. Identify the geologically vulnerable areas of salt caverns and surrounding rocks involved in the proposed adjustment of the injection-production plan.

[0037] It should be noted that geologically weak areas can be understood as areas with discontinuous geological structures, thin rock salt layers, faults or fissures, or poor mechanical properties of surrounding rocks.

[0038] S52. Based on the proposed adjustment of the injection-production plan, predict the impact of the adjusted injection-production plan on the stress state and creep rate of the salt cavern surrounding rock; S53. Based on the stress state and the creep rate, evaluate the cumulative impact of the stress state and the creep rate on the stability of the salt cavern surrounding rock structure; S54. Based on the stress state and the creep rate, evaluate the potential changes of the stress state and the creep rate on the evolution trend of fluid transport efficiency between salt caverns; S55. When the assessment results show that the proposed adjustment of the injection and production plan may cause the cumulative deformation of the surrounding rock to exceed the threshold or significantly change the long-term evolution trend of the fluid transport efficiency, a long-term risk warning is triggered. S56. Based on the aforementioned long-term risk warning, revise the injection and production plan.

[0039] It should be noted that a mechanical model of the surrounding rock of the salt cavern can be established, and parameters such as injection and production pressure and temperature can be used to simulate the stress distribution and deformation of the surrounding rock under different injection and production conditions. Creep rate can be understood as the rate at which the surrounding rock slowly deforms under long-term stress, and its prediction is crucial for assessing long-term stability. Based on this, the cumulative impact on the structural stability of the salt cavern surrounding rock is assessed according to the predicted stress state and creep rate. The cumulative impact refers to the long-term effect of stress changes and creep deformation on the structural integrity and bearing capacity of the surrounding rock throughout the entire operating cycle. Simultaneously, it is also necessary to assess the potential changes in the evolution trend of fluid transport efficiency between salt caverns caused by stress state and creep rate. This is because the deformation of the surrounding rock may alter the geometry and connectivity of the salt cavern, thereby affecting the fluid transport efficiency between salt caverns. When the assessment results show that the proposed adjustment to the injection and production plan may cause the cumulative deformation of the surrounding rock to exceed a preset threshold, or significantly change the long-term evolution trend of fluid transport efficiency, a long-term risk warning will be triggered. This threshold can be preset based on the design life, safety factor, and geological conditions of the salt cavern.

[0040] Simultaneously, it is necessary to assess the potential alterations in stress state and creep rate to the evolution trend of fluid transport efficiency between salt caverns. This is because deformation of the surrounding rock can change the geometry and connectivity of the salt caverns, thereby affecting the fluid transport efficiency between them. A long-term risk warning will be triggered when the assessment results indicate that the proposed adjustment to the injection-production plan may cause cumulative deformation of the surrounding rock to exceed a preset threshold, or significantly alter the long-term evolution trend of fluid transport efficiency. This threshold can be pre-set based on the salt cavern's design life, safety factor, and geological conditions. The purpose of the long-term risk warning is to alert the operator that the current adjustment plan may bring unacceptable long-term risks. Finally, the injection-production plan is revised based on the triggered long-term risk warning. The revised plan aims to avoid or mitigate adverse effects on the stability of the surrounding rock structure and the long-term evolution trend of fluid transport efficiency, ensuring the long-term safe and efficient operation of the salt cavern gas storage facility.

[0041] As a specific implementation method, a concrete example is given below. Suppose that during the operation of a connected salt cavern gas storage facility, due to external natural gas demand forecasts indicating a significant increase in natural gas export volume in the near future, the system initially generates a high-intensity injection-production plan. Based on the basic plan, this plan may meet demand in the short term, but the proposed solution will further evaluate it in depth. First, the system identifies geologically weak areas in the salt caverns and surrounding rocks involved in the high-intensity injection-production plan; for example, it discovers a known microfracture zone in the top surrounding rock of one of the salt caverns. Next, based on the high-intensity injection-production plan, the system predicts its impact on the stress state and creep rate of the surrounding rock of the salt cavern, particularly the microfracture zone area. Simulation results may show that under continuous high-intensity injection-production, the stress in the surrounding rock of this area will increase significantly, and the creep rate will accelerate. Subsequently, the system will evaluate the cumulative impact of these stress states and creep rates on the structural stability of the salt cavern surrounding rock, as well as the potential changes in the evolution trend of fluid transport efficiency between salt caverns. If the assessment results indicate that the cumulative deformation of the microfracture zone may exceed a preset safety threshold during the predicted operating period, or that such deformation will significantly alter the long-term fluid transport efficiency between salt caverns, such as causing local blockage or decreased connectivity, the system will immediately trigger a long-term risk warning. Based on this warning, the operator will not directly execute the initial high-intensity injection-production plan but will instead revise it. The revised plan may include: reducing the injection-production intensity of the specific salt cavern, adjusting the injection-production cycle, or introducing other salt caverns to share the burden, in order to avoid excessive stress concentration and cumulative deformation in geologically weak areas, thereby ensuring the long-term structural stability and continuous fluid transport efficiency of the salt cavern gas storage facility.

[0042] It should be noted that step S2 includes: S21. Monitor the injection / production flow rate and the pressure change rate to identify the high-intensity injection / production condition; S22. Under high-intensity injection and production conditions, the real-time pressure data and the injection and production flow data are subjected to dynamic time window smoothing processing, and the window length of the dynamic time window smoothing processing is adaptively adjusted according to the injection and production intensity and the pressure fluctuation amplitude. S23. Using the smoothed pressure data and the smoothed flow data, combined with a physical model that considers the nonlinear flow effects in the wellbore and near-wellbore zone, the current fluid transfer efficiency between the multiple salt caverns is calculated in reverse.

[0043] Specifically, monitoring the rate of change in injection and production flow rates and pressures aims to obtain key dynamic information about salt cavern operations in real time. Continuous monitoring of this data can identify high-intensity injection and production conditions, such as rapid gas injection or production, which are typically accompanied by significant changes in flow rates and pressures. The purpose of identifying high-intensity injection and production conditions is to adopt more refined data processing and model application strategies for these specific conditions to address their complexity and uncertainties.

[0044] In high-intensity injection and production operations, dynamic time-window smoothing is applied to real-time pressure and injection / production flow data. The aim is to filter out instantaneous noise and abnormal fluctuations in the data, improving data reliability. The window length for dynamic time-window smoothing can be adaptively adjusted based on the injection / production intensity and pressure fluctuation amplitude. For example, when the injection / production intensity is high or the pressure fluctuation amplitude is large, a shorter window length can be used to respond more quickly to system changes; conversely, when the injection / production intensity is low or the pressure fluctuation amplitude is small, a longer window length can be used to achieve a more stable smoothing effect. This adaptive adjustment ensures the flexibility and effectiveness of data smoothing, enabling it to better adapt to the dynamic characteristics of salt cavern operation.

[0045] In some preferred embodiments, a specific example is given below. Assume a connected salt cavern gas storage facility is undergoing high-intensity gas production during the peak winter gas consumption period. At this time, the data receiving module continuously receives real-time pressure data and injection / production flow rate data from multiple salt caverns. The efficiency calculation module first detects a significant increase in both the injection / production flow rate and pressure change rate, thus identifying the current high-intensity gas production condition. Therefore, the efficiency calculation module performs dynamic time window smoothing on the received real-time pressure data and injection / production flow rate data. Specifically, due to the high gas production intensity and drastic pressure fluctuations, the system adaptively adjusts the smoothing window length to a shorter time interval, such as 5 minutes, to quickly respond to changes in operating conditions and effectively filter out high-frequency noise. Subsequently, using the smoothed pressure and flow rate data, the efficiency calculation module invokes a pre-established physical model (e.g., a numerical model based on the finite element or finite difference method) that considers the nonlinear pressure drop effect in the wellbore and near-wellbore zone to calculate the fluid transfer efficiency between the multiple salt caverns at the current moment. In this way, even under extreme operating conditions, accurate and stable fluid transfer efficiency values ​​can be obtained, providing a reliable basis for subsequent operation scheduling and risk prediction.

[0046] Alternatively, step S3 may also include: S321. Continuously monitor the historical data of salt cavern injection and production and geological environmental parameters; S322. Based on the historical injection and production data and the geological environment parameters, identify the operational phase transition points and areas of local geological structural anomalies; S323. Based on the aforementioned operational phase transition points, the historical fluid transmission efficiency data is divided into multiple operational phases; S324. For each of the plurality of operational phases, analyze the fluid transport efficiency data within that operational phase and identify the evolution pattern within that operational phase. S325. Based on the local geological structure anomaly area, perform local enhancement analysis on the fluid transport efficiency data related to the local geological structure anomaly area to identify the unique evolution pattern of the local geological structure anomaly area. S326. Integrate the evolution patterns of each operational stage and the unique evolution patterns of the local geological structural anomaly areas to construct a segmented or multi-mode fluid transport efficiency evolution function.

[0047] Specifically, continuous monitoring of historical injection and production data and geological environmental parameters of salt caverns refers to the uninterrupted collection and recording of various data generated during the long-term operation of salt caverns. This includes, but is not limited to, operational data such as gas injection volume, gas production volume, injection and production pressure, temperature, and brine treatment volume, as well as geological environment-related parameters such as formation pressure, geostress, surrounding rock creep rate, and seismic activity. Continuous monitoring of this data provides a foundation for a comprehensive understanding of the operational status and geological response of salt caverns.

[0048] The identification of operational transition points and areas of local geological structural anomalies, based on the historical injection and production data and the geological environmental parameters, refers to identifying significant changes that may occur during the operation of the salt cavern through analysis of long-term monitoring data. Simultaneously, by combining geological environmental parameters, geological structures that may locally affect fluid transport efficiency are identified. Operational transition points can be identified using statistical methods (such as abrupt change detection in time series analysis) or based on expert experience. Areas of local geological structural anomalies can be identified through geological exploration data, geophysical monitoring data (such as microseismic monitoring), and anomalous responses in historical operational data.

[0049] In practical applications, dividing historical fluid transport efficiency data into multiple operational stages based on the aforementioned operational stage transition points means that once an operational stage transition point is identified, the system will automatically or manually divide the historical fluid transport efficiency data for the entire operational cycle into several independent operational stages. Each operational stage represents a relatively stable or characteristic operating state of the salt cavern under specific operating conditions and geological background. The purpose is to independently analyze the fluid transport characteristics under different operational stages to avoid interference from the complexities of different stages.

[0050] Furthermore, for each of the multiple operational phases, analyzing the fluid transport efficiency data within that operational phase and identifying the evolutionary pattern within that phase refers to conducting in-depth analysis of the fluid transport efficiency data within each defined operational phase. This can include trend analysis, periodic analysis, correlation analysis, etc., to reveal the specific patterns or regularities of fluid transport efficiency changes over time within that phase.

[0051] Furthermore, based on the local geological structural anomaly area, local enhancement analysis is performed on the fluid transport efficiency data related to the local geological structural anomaly area. Identifying the unique evolution pattern of the local geological structural anomaly area refers to a more detailed and focused analysis of the fluid transport efficiency data of the nearby or affected salt caverns for the identified local geological structural anomaly area.

[0052] Therefore, integrating the evolutionary patterns of each operational stage and the unique evolutionary patterns of the local geological anomaly areas to construct a piecewise or multi-mode fluid transport efficiency evolution function refers to synthesizing the evolutionary patterns identified from different operational stages and local geological anomaly areas. This can be done by constructing a piecewise function, where each segment corresponds to the evolutionary pattern of an operational stage or anomaly area; or by constructing a multi-mode function that can dynamically select or combine different evolutionary patterns for prediction based on the current operational status and geological conditions. The aim is to establish a mathematical model that can comprehensively and accurately reflect the complex evolutionary laws of salt cavern fluid transport efficiency.

[0053] In some preferred embodiments, a specific example is given below. Assume a connected salt cavern gas storage facility has undergone three main operational phases over the past decade: an initial injection and construction phase (years 1-3), a stable high-intensity injection and production phase (years 4-8), and a recent low-intensity peak-shaving phase (years 9-10). Meanwhile, geological exploration data and microseismic monitoring indicate a localized fault zone in the surrounding rock of one of the salt caverns, which may affect fluid transport.

[0054] First, the system continuously monitors historical data such as injection and production flow rate, pressure, and temperature of the salt caverns over the past decade, as well as geological environmental parameters such as geostress and creep rate.

[0055] Secondly, based on this data, the operational transition points at the end of year 3 and year 8 were identified. Simultaneously, this local fault zone was identified as a region of geological structural anomalies.

[0056] Next, the historical fluid transport efficiency data over the past decade was divided into three operational phases. For each operational phase, such as the stable high-intensity injection and production phase, the fluid transport efficiency data was analyzed, revealing that the efficiency in this phase showed a slow downward trend, accompanied by seasonal fluctuations.

[0057] Meanwhile, for this local fracture zone, a local enhancement analysis was performed on the fluid transport efficiency data related to this area. It was found that the fluid transport efficiency in this area would suddenly and nonlinearly decrease during certain high-pressure gas injection periods, which is related to the stress response of the fracture zone.

[0058] Finally, the system integrates the evolution patterns of these three operational phases (e.g., rapid stabilization in the initial phase, linear decline with seasonal fluctuations in the stable phase, and smooth fluctuations in the low-intensity phase) with the unique evolution patterns of local fault zones (e.g., nonlinear decline under high pressure) to construct a piecewise or multi-mode fluid transport efficiency evolution function. This function can dynamically select or combine corresponding evolution patterns for prediction based on the current operational phase and whether it is affected by local fault zones, thus providing a more accurate and comprehensive fluid transport efficiency evolution model.

[0059] Step S4 includes: Step S4 includes: S41. Under normal operating conditions, simplified scenario simulation is used to extrapolate future operating scenarios based on macro parameters and to make preliminary risk predictions. S42. When the preliminary risk assessment results show that there is a potential high risk, or when the rate of change of injection or production flow or pressure exceeds the preset threshold, a refined scenario simulation is triggered. S43. In refined scenario simulation, for high-risk areas or specific salt caverns, local geological structure parameters and detailed fluid transport efficiency evolution functions are introduced to conduct high-resolution simulations, and risk predictions are made based on the refined simulation results.

[0060] In some preferred embodiments, this application is implemented as follows. Assume a connected salt cavern gas storage facility is undergoing routine gas injection operations. The system first employs a simplified scenario simulation under normal operating conditions. This simulation rapidly extrapolates the operating scenario for the next 72 hours based on macroscopic parameters such as the average injection flow rate over the past 24 hours, the average pressure of the salt cavern cluster, and the natural gas market demand forecast for the next week. Preliminary risk assessment results show that, at the current injection rate, the overall pressure change trend of the salt cavern cluster is stable, with no significant anomalies.

[0061] However, during a certain gas injection process, the system detected that the gas injection flow rate of a specific salt cavern 101 suddenly exceeded the preset threshold, and its pressure change rate also accelerated significantly. At this point, the system immediately triggers a refined scenario simulation. In the refined scenario simulation, a detailed three-dimensional geological model of salt cavern 101, the creep parameters of the surrounding rock, and the fluid transport efficiency evolution function specific to this region are introduced for salt cavern 101 and its adjacent areas. Through high-resolution numerical simulation, the system can accurately predict the local stress distribution, creep rate, and potential impact on the fluid transport efficiency of adjacent salt caverns in the surrounding rock of salt cavern 101. The refined simulation results show that if the current high-intensity gas injection is maintained, the surrounding rock of salt cavern 101 may experience local cumulative deformation exceeding the safety threshold within the next 48 hours. Based on this refined risk prediction, the system will trigger a long-term risk warning and recommend immediate adjustments to the gas injection plan for salt cavern 101, such as reducing the gas injection rate or suspending gas injection, to avoid potential structural risks.

[0062] Step S43 includes: S431. Obtain the distribution range and confidence level information of local geological structure parameters; S432. Obtain the fluctuation range and correlation information of key parameters in the fluid transport efficiency evolution function; S433. Based on the distribution range of the local geological structure parameters, the confidence information of the local geological structure parameters, the fluctuation range of the key parameters in the fluid transport efficiency evolution function, and the correlation information of the key parameters in the fluid transport efficiency evolution function, generate multiple sets of simulation parameter combinations. S434. Perform high-resolution simulation for each set of simulation parameter combinations to obtain the corresponding simulation results; Statistical analysis was performed on the multiple sets of simulation results to obtain the confidence intervals of the risk prediction results.

[0063] Specifically, local geological structural parameters can be understood as parameters describing the physical properties of salt caverns and their surrounding rocks, such as rock porosity, permeability, elastic modulus, Poisson's ratio, compressive strength, and creep parameters. The distribution range of these parameters can be obtained from various sources, including geological exploration data, well logging data, core analysis, and historical operational data. Their confidence level reflects the measurement accuracy, data coverage, and reliability of model assumptions. For example, regional geological parameters obtained through seismic exploration may have a large distribution range and low confidence level, while local parameters obtained through borehole core testing may have a smaller distribution range and higher confidence level.

[0064] The fluctuation range and correlation information of key parameters in the fluid transport efficiency evolution function refer to the core variables in the function model that affect the time-varying fluid transport efficiency between salt caverns. These key parameters may include the evolution rate of salt cavern wall roughness, fracture propagation rate, and the sensitivity coefficient of the salt rock creep rate to the influence of the transport channel. Their fluctuation range can be estimated through long-term operational data analysis, laboratory simulations, and expert experience, while the correlation information describes the degree and pattern of interaction between different key parameters. For example, an increase in the salt rock creep rate may be positively correlated with an acceleration of the fracture propagation rate.

[0065] In practical applications, multiple sets of simulation parameter combinations are generated based on the distribution range of the local geological structure parameters, the confidence level of the local geological structure parameters, the fluctuation range of key parameters in the fluid transport efficiency evolution function, and the correlation information of the key parameters in the fluid transport efficiency evolution function. The purpose is to construct a series of input parameter sets that can represent actual uncertainties. For example, Monte Carlo simulation, Latin hypercube sampling, or other random sampling methods can be used to extract values ​​from the distribution range of each parameter and combine them with their confidence level and correlation information to generate thousands or even tens of thousands of unique parameter combinations. Each parameter combination represents a possible future scenario.

[0066] Therefore, for each set of simulation parameters, a high-resolution simulation is performed to obtain the corresponding simulation results. This means that for each set of parameters generated above, a detailed high-resolution numerical simulation considering the evolution of local geological structure and fluid transport efficiency will be run to predict the operational performance of the salt cavern under that specific parameter combination, such as pressure changes, injection-production flow rates, surrounding rock stress distribution, deformation, and fluid transport efficiency.

[0067] Furthermore, statistical analysis is performed on the multiple sets of simulation results to obtain the confidence intervals for the risk prediction results. Specifically, after obtaining a large number of simulation results, key risk indicators (e.g., the probability that salt cavern deformation exceeds the safety threshold, the time it takes for fluid transport efficiency to drop below the critical value, etc.) can be statistically processed. For example, the mean, standard deviation, and median of these indicators can be calculated, and their probability density functions can be constructed. By analyzing these statistics, the confidence intervals for the risk prediction results can be determined, such as 90% or 95% confidence intervals, thereby quantifying the uncertainty of risk prediction and providing decision-makers with more comprehensive risk assessment information.

[0068] Additionally, step S434 includes: S43411. Conduct a preliminary assessment of the key risk indicators in the multiple sets of simulation results and identify outliers that exceed the normal range; S43412. Combining the known risk event characteristics and confidence information of geological structure parameters in historical operational data, the physical rationality of the outlier is verified. S43413. When the outlier is determined to be physically reasonable, a dynamically adjusted weight is applied to the outlier. The weight is adjusted based on the degree of matching between the outlier and the characteristics of the historical risk event and the confidence level of the geological structure parameters. S43414. When the outlier is determined to be physically unreasonable, the outlier shall be removed or corrected. S43415. Based on the processed multiple sets of simulation results, calculate the confidence interval of the risk prediction result.

[0069] Specifically, a preliminary assessment of key risk indicators in the aforementioned sets of simulation results is conducted to identify data points that significantly deviate from the expected distribution or normal range. These key risk indicators may include the stress state of the salt cavern surrounding rock, creep rate, short-term fluctuations in fluid transport efficiency, or long-term evolution trends. Outliers can be identified using statistical methods, such as those based on standard deviation, quartile range (IQR), or machine learning anomaly detection algorithms.

[0070] The process of verifying the physical plausibility of outliers by combining known risk event characteristics and confidence information of geological structural parameters from historical operational data involves in-depth analysis of identified outliers to determine whether they conform to actual physical processes or geological conditions. Historical operational data can provide patterns and characteristics of past risk events, such as thresholds for surrounding rock deformation under specific injection and production conditions, and precursors to abnormal changes in fluid transport efficiency. Confidence information of geological structural parameters reflects the level of understanding of the accuracy of local geological models. For example, in areas with sparse geological exploration data, the confidence level of geological parameters may be low, leading to increased uncertainty in simulation results. By comparing outliers with this information, it can be determined whether the outlier is a genuine potential risk signal or an artifact caused by model or data input errors.

[0071] In practical applications, when the aforementioned outliers are deemed physically plausible, dynamically adjusted weights are applied to them. These weights are adjusted based on the degree of matching between the outlier and the characteristics of the historical risk events, as well as the confidence level of the geological structural parameters. This means that simulation results that, while anomalous, are considered to accurately reflect potential risks should not be simply discarded but rather assigned appropriate weights. For example, if an outlier highly matches the characteristics of a major historical risk event and its corresponding geological structural parameters have a high confidence level, it can be assigned a higher weight to ensure it is fully considered in the final risk assessment. Conversely, if the matching degree is low or the confidence level of the geological parameters is not high, the weight can be adjusted accordingly.

[0072] Furthermore, when the aforementioned outliers are determined to lack physical plausibility, they are either removed or corrected. Outliers lacking physical plausibility are typically caused by model calculation errors, incorrect input data, or extremely unrealistic parameter combinations. For these outliers, they can be directly removed from the simulation results set, or corrected using interpolation, regression, or other methods to restore them to a physically plausible range, thus avoiding any negative impact on the final confidence interval calculation.

[0073] Therefore, based on the processed simulation results, the confidence intervals for the risk predictions are calculated. After effectively handling outliers, the remaining simulation results will more accurately reflect the system's true behavior and uncertainties. At this point, traditional statistical methods, such as Monte Carlo simulation, Bootstrap resampling, or parameter estimation, can be used to calculate more representative and reliable confidence intervals for the risk predictions.

[0074] Additionally, step S434 may also include: S43421. Identify the nonlinear correlations among key risk indicators in the multiple sets of simulation results; S43422. Construct a multidimensional risk space based on the identified nonlinear correlations; S43423. In the multidimensional risk space, perform distribution feature mapping on the multiple sets of simulation results; S42424. In the multidimensional risk space, high-risk clustered areas and low-risk sparse areas are identified according to the distribution feature mapping. S43425. Determine the risk boundary based on the identified high-risk clustered areas and low-risk sparse areas; S43426. Based on the determined risk boundary, calculate the confidence interval of the risk prediction result.

[0075] Specifically, identifying the nonlinear correlations among key risk indicators in the multiple sets of simulation results refers to using advanced data analysis techniques, such as machine learning algorithms (e.g., support vector machines, neural networks, decision trees) or nonlinear regression models, to conduct in-depth analysis of multiple key risk indicators involved in the simulation results generated by the refined scenario simulation. These key risk indicators may include, but are not limited to, salt cavern surrounding rock stress, creep rate, fluid transport efficiency change rate, gas storage fluctuation, and formation pressure response. The aim is to reveal the nonlinear dependencies among these indicators, such as threshold effects, saturation effects, hysteresis effects, or complex interactions, rather than simple linear proportional relationships.

[0076] The construction of a multidimensional risk space based on the identified nonlinear correlations can be understood as treating each key risk indicator as an independent dimension of this space, and using the previously identified nonlinear correlations to define the interaction rules and structure between these dimensions. This results in a multidimensional mathematical space capable of comprehensively and precisely characterizing the operational risk status of salt caverns. In this multidimensional risk space, each set of simulation results can be mapped to a unique point, whose coordinate values ​​in each dimension are determined by the corresponding key risk indicator values ​​in that set of simulation results.

[0077] In practical applications, mapping the distribution characteristics of the multiple sets of simulation results within the multidimensional risk space specifically involves using methods such as cluster analysis, density estimation (e.g., kernel density estimation), manifold learning, or dimensionality reduction visualization to visualize or quantify the distribution patterns of the multiple sets of simulation results in the multidimensional risk space. This helps to intuitively reveal the central tendency, dispersion, potential structural characteristics, and outliers of the simulation results across different risk dimensions, such as the existence of multiple risk centers, specific risk paths, or irregular risk regions.

[0078] Furthermore, in the multidimensional risk space, identifying high-risk clustered areas and low-risk sparse areas based on the distribution feature mapping means, through in-depth analysis of the density distribution of simulation results in the multidimensional risk space, determining those areas where simulation result points are highly concentrated as high-risk clustered areas. These areas represent scenario combinations where the probability of risk events occurring is high or the degree of risk is relatively severe under specific operating conditions. Conversely, areas where simulation result points are sparsely distributed or almost non-existent are identified as low-risk sparse areas, which typically correspond to scenarios with lower risk or more stable operating conditions.

[0079] Therefore, determining risk boundaries based on the identified high-risk clustered areas and low-risk sparse areas means using classification algorithms (such as support vector machines, decision trees, and random forests) or geometric methods (such as convex hull algorithms and contour analysis) to delineate one or more boundaries that can clearly distinguish regions with different risk levels, based on the distribution characteristics of these areas. These risk boundaries can be non-linear to better adapt to the complexity and non-linear characteristics of the multidimensional risk space, thereby more accurately defining risk areas.

[0080] Finally, calculating the confidence interval for the risk prediction result based on the determined risk boundary involves quantifying the range of uncertainty in the risk prediction result by analyzing the distribution of multiple sets of simulation result points relative to the determined risk boundary. For example, the proportion of simulation results falling within a specific risk boundary (e.g., defined as a high-risk area) can be calculated, or a probability distribution of different risk levels can be defined based on the boundary, thereby obtaining a more robust and accurate confidence interval. This confidence interval reflects the potential risk range under full consideration of nonlinear factors and multidimensional interactions, providing a more reliable basis for decision-making.

[0081] A system for operating and scheduling interconnected salt cavern gas storage facilities, comprising: The data receiving module is used to receive real-time pressure data and injection / production flow data from multiple salt caverns; the efficiency calculation module is used to reverse-calculate the current fluid transfer efficiency between the multiple salt caverns based on the real-time pressure data and the injection / production flow data. An evolution function establishment module is used to track the trajectory of the fluid transport efficiency changing over time, and to identify the evolution trend of the fluid transport efficiency based on the trajectory, thereby establishing the evolution function of the fluid transport efficiency; The scenario simulation and risk prediction module is used to simulate future operating scenarios based on the current fluid transmission efficiency, the evolution function of the fluid transmission efficiency, and external natural gas demand forecasts, and to predict potential operating risks based on the simulation results. The plan adjustment module is used to trigger an early warning when the potential operational risks are anticipated, and to adjust the injection and production plan according to the potential operational risks.

[0082] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0083] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for scheduling operation of a connected salt cavern gas storage, characterized in that, The method comprises the following steps: S1, receiving real-time pressure data and injection-production flow rate data from multiple salt caverns; S2, inversely calculating the current fluid transfer efficiency between the multiple salt caverns according to the real-time pressure data and the injection-production flow rate data; S3, tracking the trajectory of the fluid transfer efficiency over time, and identifying the evolution trend of the fluid transfer efficiency according to the trajectory, thereby establishing an evolution function of the fluid transfer efficiency; S4, based on the current fluid transfer efficiency, the evolution function of the fluid transfer efficiency, and external natural gas demand prediction, simulating future operation scenarios, and predicting potential operation risks according to the simulation results; S5, when the potential operation risks are predicted, triggering an early warning, and adjusting the injection-production plan according to the potential operation risks.

2. The method according to claim 1, wherein, Step S3 comprises: S311, receiving real-time pressure data, injection-production flow rate data, and auxiliary data related to local geologic stress from multiple salt caverns; S312, simultaneously optimizing the fluid transfer efficiency parameter and the sensor drift parameter according to the real-time pressure data and the injection-production flow rate data; S313, adjusting the preference for changes in the fluid transfer efficiency parameter and the sensor drift parameter according to the auxiliary data; S314, assigning a confidence weight to the fluid transfer efficiency parameter; S315, learning the evolution function from the confidence-weighted fluid transfer efficiency parameter; S316, triggering a deviation-driven reassessment when there is a persistent deviation between the trend predicted by the evolution function and the current high-confidence fluid transfer efficiency parameter.

3. The method according to claim 1, wherein, Step S5 comprises: S51, identifying the geological weak areas of the salt caverns and their surrounding rocks involved in the injection-production plan to be adjusted; S52, predicting the impact of the injection-production plan to be adjusted on the stress state and creep rate of the salt cavern surrounding rocks according to the injection-production plan to be adjusted; S53, evaluating the cumulative impact of the stress state and the creep rate on the structural stability of the salt cavern surrounding rocks according to the stress state and the creep rate; S54, evaluating the potential changes in the evolution trend of the fluid transfer efficiency between salt caverns according to the stress state and the creep rate; S55, triggering a long-term risk warning when the evaluation results show that the injection-production plan to be adjusted may cause the cumulative deformation of the surrounding rocks to exceed a threshold or significantly change the long-term evolution trend of the fluid transfer efficiency; S56, revising the injection-production plan according to the long-term risk warning. Step S2 comprises:

4. The method according to claim 1, wherein, S21, monitoring the injection-production flow rate and the pressure change rate to identify high-intensity injection-production working conditions; S22, under high-intensity injection-production working conditions, performing dynamic time window smoothing processing on the real-time pressure data and the injection-production flow rate data, and the window length of the dynamic time window smoothing processing is adaptively adjusted according to the injection intensity and the pressure fluctuation amplitude; S23, inversely calculating the current fluid transfer efficiency between the multiple salt caverns using the smoothed pressure data and the smoothed flow rate data, and considering the physical model of nonlinear flow effects in the wellbore and near-wellbore zone. Step S3 comprises:

5. The method according to claim 1, wherein, ​ S321, continuously monitor injection-production history data and geological environment parameters of the salt cavern; S322, identify an operation stage transition point and a local geological structure abnormal area according to the injection-production history data and the geological environment parameters; S323, divide historical fluid transmission efficiency data into multiple operation stages according to the operation stage transition point; S324, for each operation stage of the multiple operation stages, analyze fluid transmission efficiency data in the operation stage to identify an evolution mode in the operation stage; S325, according to the local geological structure abnormal area, perform local enhancement analysis on fluid transmission efficiency data related to the local geological structure abnormal area to identify a unique evolution mode of the local geological structure abnormal area; S326, integrate the evolution modes of the operation stages and the unique evolution modes of the local geological structure abnormal areas to construct a segmented or multi-mode fluid transmission efficiency evolution function.

6. The method according to claim 1, wherein, Step S4 includes: S41, in a normal operation state, a simplified scenario simulation is used to deduce future operation scenarios based on macro parameters and perform preliminary risk prediction; S42, when the preliminary risk prediction result shows that there is a potential high risk, or when the injection-production flow rate or pressure change rate exceeds the preset threshold, trigger a refined scenario simulation; S43, in the refined scenario simulation, local geological structure parameters and detailed fluid transmission efficiency evolution functions are introduced for high-risk areas or specific salt caverns, high-resolution simulation is performed, and risk prediction is performed according to the refined simulation result.

7. The method according to claim 6, wherein, Step S43 includes: S431, obtain the distribution range and confidence information of the local geological structure parameters; S432, obtain the fluctuation range and correlation information of the key parameters in the fluid transmission efficiency evolution function; S433, according to the distribution range of the local geological structure parameters, the confidence information of the local geological structure parameters, the fluctuation range of the key parameters in the fluid transmission efficiency evolution function, and the correlation information of the key parameters in the fluid transmission efficiency evolution function, generate multiple sets of simulation parameter combinations; S434, for each set of simulation parameter combinations, perform high-resolution simulation to obtain the corresponding simulation result; statistical analysis is performed on the multiple simulation results to obtain a confidence interval of the risk prediction result.

8. The method according to claim 7, wherein, Step S434 includes: S43411, preliminarily evaluate key risk indicators in the multiple simulation results to identify abnormal values that exceed the normal range; S43412, combined with known risk event characteristics in historical operation data and confidence information of geological structure parameters, physically verify the abnormal values; S43413, when the abnormal values are judged to have physical rationality, a dynamically adjusted weight is applied to the abnormal values, and the weight is adjusted according to the matching degree of the abnormal values with the historical risk event characteristics and the confidence of the geological structure parameters; S43414, when the abnormal values are judged to be not physically reasonable, the abnormal values are removed or corrected; S43415. Based on the processed multiple sets of simulation results, calculate the confidence interval of the risk prediction result.

9. The method according to claim 7, wherein, Step S434 includes: S43421. Identify the nonlinear correlations among key risk indicators in the multiple sets of simulation results; S43422. Construct a multidimensional risk space based on the identified nonlinear correlations; S43423. In the multidimensional risk space, perform distribution feature mapping on the multiple sets of simulation results; S42424. In the multidimensional risk space, high-risk clustered areas and low-risk sparse areas are identified according to the distribution feature mapping. S43425. Determine the risk boundary based on the identified high-risk clustered areas and low-risk sparse areas; S43426. Based on the determined risk boundary, calculate the confidence interval of the risk prediction result.

10. A communication system for operating and scheduling a salt cavern gas storage, characterized in that, The system includes: The data receiving module is used to receive real-time pressure data and injection / production flow data from multiple salt caverns; the efficiency calculation module is used to reverse-calculate the current fluid transfer efficiency between the multiple salt caverns based on the real-time pressure data and the injection / production flow data. An evolution function establishment module is used to track the trajectory of the fluid transport efficiency changing over time, and to identify the evolution trend of the fluid transport efficiency based on the trajectory, thereby establishing the evolution function of the fluid transport efficiency; The scenario simulation and risk prediction module is used to simulate future operating scenarios based on the current fluid transmission efficiency, the evolution function of the fluid transmission efficiency, and external natural gas demand forecasts, and to predict potential operating risks based on the simulation results. The plan adjustment module is used to trigger an early warning when the potential operational risks are anticipated, and to adjust the injection and production plan according to the potential operational risks.

Images