A multi-dimensional signal processing method, calculation method and system for shale gas well operation and maintenance evaluation

Abstract

This invention belongs to, but is not limited to, the field of shale gas production technology, and particularly relates to a multi-dimensional health and risk scoring calculation method and system for operation and maintenance assessment, including: S1. Index preprocessing: Standardizing the input multi-dimensional indicators to ensure the accuracy of subsequent calculations; S2. Multi-dimensional aggregation: Based on the preprocessed indicators, aggregating them into dual core indicators of health and risk through a weighted scoring model; S3. Score correction: Eliminating single-well data bias through two correction methods: cross-well comparison and operating condition identification, ensuring that the score fits the actual operation and maintenance scenario of shale gas wells; S4. Status output: Mapping the corrected health-risk dual indicators into a three-stage production status.

Description

Technical Field

[0001] This invention belongs to, but is not limited to, the field of shale gas production technology, and particularly relates to a multi-dimensional signal processing method, calculation method, and system for shale gas well operation and maintenance assessment. Background Technology

[0002] Shale gas, as an important unconventional natural gas resource, has been developed and utilized on a large scale in recent years. With increasing production years, shale gas wells in most areas of China are gradually entering the aging stage, resulting in fluctuating and unstable production capacity. According to statistics for the entire year of 2024, the Southwest Oil and Gas Field had an average of 10 wells experiencing liquid accumulation per day, with an average well recovery cycle of 2.5 days and an average daily production of 15,000 cubic meters. This resulted in a daily production loss of up to 150,000 cubic meters, impacting annual production by approximately 55 million cubic meters. Improving the response efficiency to abnormal conditions such as liquid accumulation and enhancing the intelligence level of production operation and maintenance have become urgent problems to be solved in current shale gas production and operation.

[0003] In existing technologies, the identification and handling of anomalies such as wellbore fluid accumulation, compressor shutdown, and pipeline blockage during shale gas production still mainly rely on manual experience and static threshold alarm mechanisms set in SCADA (Supervisory Control and Data Acquisition) or POC (Production Operation Control) systems. This type of method has the following significant drawbacks: 1. Fixed rules and poor generalization ability: The SCADA / POC system alarms based on manually set rules, which cannot adapt to the complex working conditions brought about by the evolution of wellbore status over time and environmental changes, and is prone to missed alarms or false alarms.

[0004] 2. Reliance on manual analysis and delayed response: Current anomaly assessment still relies on the experience and manual analysis of on-site engineers, which is inefficient and makes it difficult to achieve accurate and rapid dynamic management.

[0005] 3. Low data utilization: Although downhole sensors and surface monitoring equipment have achieved the acquisition of a large amount of time-series data, the existing system has failed to fully explore the dynamic evolution patterns hidden within them.

[0006] 4. Lack of model capabilities: There is a lack of intelligent models with self-learning and automatic feature extraction capabilities, making it difficult to establish a detection system with high robustness and adaptability in multi-well and multi-dimensional data environments.

[0007] Based on the above analysis, the urgent technical problems that need to be solved in the existing technology are: the existing general equipment health rating scheme, in the shale gas well operation and maintenance scenario, does not consider the coupling characteristics of operating conditions, relies on a single indicator / fixed threshold for judgment, lacks the cumulative impact of historical data and cross-well comparison correction, and the output results are difficult to directly connect with operation and maintenance decisions, resulting in scoring adaptation bias, insufficient accuracy, delayed risk assessment and low operation and maintenance practicality. Summary of the Invention

[0008] To address the problems existing in the prior art, this invention provides a method and system for calculating multidimensional health and risk scores for operation and maintenance assessment.

[0009] This invention is implemented as follows: a multi-dimensional signal processing method for shale gas well operation and maintenance assessment, comprising the following steps: The acquired multi-dimensional operation and maintenance signals are time-aligned so that continuous and discrete signals at different time scales are uniformly mapped to the same statistical period.

[0010] Anomaly suppression processing is performed on maintenance signals that have completed time alignment to reduce the impact of transient disturbances on signal trends.

[0011] Based on the preset index grading boundaries, the processed continuous signal is converted into a graded score signal.

[0012] When calculating the graded score for the current period, historical memory weights are introduced to weight and sum the scores of similar signals within historical periods. The further away the historical period is from the current period, the smaller its weight.

[0013] Furthermore, the time alignment process includes: Continuous signals are sampled and statistically analyzed according to a fixed time window, and discrete event signals are converted into frequency or duration within the time window.

[0014] Furthermore, the anomaly suppression process includes performing sliding smoothing on continuous signals and setting an upper limit truncation rule for discrete signals.

[0015] This invention also provides a method for calculating multidimensional health and risk scores for operation and maintenance assessment, including: S1. Indicator Preprocessing: Standardize the input multi-dimensional indicators to ensure the accuracy of subsequent calculations.

[0016] S2. Multi-dimensional aggregation: Based on the preprocessed indicators, they are aggregated into dual core indicators of health and risk through a weighted scoring model.

[0017] S3. Scoring Correction: Through two correction methods, cross-well comparison and operating condition identification, the bias of single-well data is eliminated to ensure that the scoring is consistent with the actual operation and maintenance scenario of shale gas wells.

[0018] S4. Status Output: Based on the corrected health and risk dual indicators, it is mapped to a three-stage production status.

[0019] Furthermore, the input data covers the entire operation and maintenance scenario of shale gas wells, providing a basis for scoring calculations, specifically including the following four categories: Real-time operating parameters: real-time gas well production and process parameters.

[0020] Historical operation and maintenance data: historical output data, historical alarm records, and historical operation event records.

[0021] Abnormal monitoring data: frequency of abnormal alarms, abnormal data on pressure difference, and probability data of fluid accumulation.

[0022] Decision support data: basic information on similar wells and key weighting of gas field operation and maintenance.

[0023] Furthermore, S1 specifically includes: setting the boundary of indicator bins: based on the operation and maintenance experience of shale gas wells and industry standards, setting the boundary of bins for indicators such as liquid accumulation probability, valve pressure fluctuation, pressure difference abnormality, alarm recovery time, and operation events, and converting continuous indicators into quantifiable hierarchical data.

[0024] Historical memory weighting is introduced: each indicator is assigned a historical memory weight. When calculating the current indicator score, the influence of indicator data within the historical period is superimposed. The weight decays over time (recent data has a higher weight), reflecting the cumulative effect of long-term operation and maintenance data on the current score.

[0025] Furthermore, S2 specifically includes: Sub-dimensional scoring calculation: Weighted scoring algorithms are designed for the three core operation and maintenance dimensions of health, liquid accumulation risk, and production potential. The health dimension is based on the deviation of production trend, plus scores for indicators such as valve post-pressure fluctuation and alarm recovery time; the risk dimension uses liquid accumulation probability, pressure difference abnormality, and abnormal alarm frequency as core indicators, plus scores for the impact of operational events.

[0026] Dual-indicator aggregation: Based on the key points of gas field operation and maintenance, the weights of sub-dimensions such as health, liquid accumulation risk, and production potential are dynamically adjusted, and the final health score and risk score are obtained by weighted summation.

[0027] Furthermore, S3 specifically includes: cross-well comparison baseline correction: establish a comparison baseline for similar wells with the same process and production cycle, compare the current well's health and risk dual indicators with the baseline data, and if the deviation exceeds the preset threshold, correct the current score according to the baseline ratio to avoid score distortion caused by special working conditions of a single well.

[0028] Operating condition pattern recognition and correction: Based on the prediction results of the liquid accumulation early warning system, the current operating condition of the gas well is automatically identified, and the operating condition results are used as a correction factor for the risk score, reflecting the direct impact of the operating condition coupling characteristics on risk.

[0029] Furthermore, S4 specifically includes: Stable state: Health score ≥ 80 points, risk score ≤ 30 points, no serious abnormal indicators, and no need for emergency operation and maintenance intervention.

[0030] Warning status: Health score of 60-80 points, or risk score of 30-60 points, with 1-2 medium-risk indicators present, requiring close monitoring and preparation of operation and maintenance plans.

[0031] Critical condition: Health score < 60 points, or risk score > 60 points, indicating critical abnormalities such as severe fluid accumulation or high pressure difference, requiring immediate operation and maintenance to be initiated.

[0032] Another objective of this invention is to provide a system for calculating multidimensional health and risk scores for operation and maintenance assessment, which implements the aforementioned method for calculating multidimensional health and risk scores for operation and maintenance assessment. The system includes: Data layer: Responsible for the collection, storage and cleaning of input data, supporting real-time data access and historical data archiving, and connecting to similar well baseline databases.

[0033] Computation layer: Deploys core computing modules, including an index bucketing engine, a historical weight calculation engine, a weighted scoring model, a cross-well comparison engine, and a working condition identification model, to achieve fully automated calculation of the preprocessing-aggregation-correction process.

[0034] Output layer: Provides a visual output interface that displays the values ​​of health and risk indicators, three-stage production status, details of abnormal indicators, and operation and maintenance suggestions. It also supports data export and integration with operation and maintenance systems.

[0035] Another object of the present invention is to provide a computer device, the computer device including a memory and a processor, the memory storing a computer program, the computer program being executed by the processor causing the processor to perform the steps of the multidimensional health and risk score calculation method for operation and maintenance assessment.

[0036] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the multidimensional health and risk score calculation method for operation and maintenance assessment.

[0037] Another objective of this invention is to provide an information data processing terminal, which includes the aforementioned multi-dimensional health and risk scoring calculation system for operation and maintenance assessment.

[0038] Based on the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solution to be protected by this invention are as follows: The core advantages of this technical solution revolve around the key pain points of shale gas well operation and maintenance, namely, the complexity of operating conditions, the multidimensionality of data, and the precision of requirements. From four dimensions—scenario adaptability, scoring accuracy, risk prediction, and operational practicality—it addresses the problems of insufficient adaptability, delayed judgment, and lack of accuracy of traditional general-purpose equipment health scoring schemes in shale gas well scenarios. A detailed analysis follows: (1) It fits the coupling characteristics of shale gas well operating conditions and eliminates the adaptation deviation of general solutions.

[0039] Traditional health assessment schemes are mostly designed for general industrial equipment and do not consider the special operating conditions of shale gas wells where multiple parameters are interconnected (such as fluid accumulation directly aggravating pressure differentials and short-term operational events temporarily interfering with pressure indicators), which can easily lead to a disconnect between the assessment and actual risk. This invention addresses this problem specifically through two design features.

[0040] The core algorithm incorporates a coupling logic specific to shale gas wells, explicitly using the short-term impact of associated operational events such as fluid accumulation and pressure differential anomalies on risk as a key factor in the scoring calculation, rather than calculating a single indicator in isolation.

[0041] The operating condition pattern recognition module (such as the initial decline in production potential during stable production liquid accumulation) is designed specifically for shale gas well operation and maintenance scenarios. It can accurately distinguish between short-term operational interference and long-term health hazards (such as the pressure fluctuation after operation in well A2 in Example 1 being judged as temporary interference, avoiding misjudgment as risk), ensuring that the scoring fits the actual operation and maintenance scenario of shale gas wells.

[0042] (2) Multi-dimensional aggregation + dynamic correction mechanism greatly improves the accuracy of scoring.

[0043] Traditional methods often rely on single-indicator threshold judgments (such as only checking whether pressure exceeds the limit) or fixed-weight calculations, which are prone to scoring distortion due to incomplete data, failure to consider historical cumulative effects, and individual well differences. This invention constructs a closed loop through a three-layer precision design.

[0044] Multi-dimensional indicator aggregation: covering real-time parameters (pressure, production) + historical data (operation and maintenance records) + abnormal data (alarms, fluid accumulation), avoiding the one-sidedness of a single indicator. For example, the health score is combined with production trend, pressure fluctuation and alarm recovery time to more comprehensively reflect the well condition.

[0045] Historical memory weight correction: Introduce historical weights with time decay (recent data has a greater impact) to reflect the cumulative effect of long-term health status (such as capturing the health decline trend of well B3 in the past six months in Example 2, rather than just looking at the current data), and avoid short-term data from masking long-term problems.

[0046] Cross-well comparison baseline correction: Establish a baseline for wells of the same type with the same process and production cycle to eliminate scoring deviations caused by special operating conditions of individual wells (such as differences in equipment aging in individual wells) and ensure that the scores are comparable among wells of the same type (such as the risk score of well A1 in Example 1 being verified by the baseline of wells of the same type, confirming that it is a real risk rather than an individual deviation).

[0047] (3) From passive threshold triggering to proactive trend prediction, avoid hidden operation and maintenance risks in advance.

[0048] Traditional solutions often employ a passive approach, only issuing alerts when indicators fall below thresholds. This approach is slow to react to common shale gas well degradation issues (such as slow production decline and subtle deterioration in well health), easily missing the optimal intervention window. This invention achieves proactive prevention through two forward-looking designs.

[0049] Production trend deviation analysis: The Arps decline model (a commonly used production prediction model for shale gas fields) is introduced to calculate the deviation between the actual production and the prediction curve. This can capture hidden problems that have deviated from the normal trend but have not fallen below the threshold (for example, in Example 2, although the production of well B3 did not reach the emergency threshold, it was detected in time that it deviated from the prediction curve by 7.69%).

[0050] Health trend tracking: Combining historical memory weights, it can dynamically track the long-term trend of health changes (instead of just looking at the current score). For example, the slow decline in the health of well B3 over the past six months is superimposed and calculated, and finally judged as a stable-early warning transition state. This prompts the operation and maintenance team to adjust parameters in advance to avoid further decline in production and realize the transformation from passively dealing with faults to actively preventing potential risks.

[0051] (4) The output results are directly linked to operation and maintenance decisions, which improves the practicality of technology implementation.

[0052] Traditional solutions typically only output a health score, requiring secondary interpretation by operations and maintenance personnel. This makes it difficult to directly guide actual operations and fails to consider the resource allocation needs of gas field operations and maintenance. This invention optimizes the output and value from a practical perspective.

[0053] Three-stage status + precise operation and maintenance suggestions: The abstract health-risk score is transformed into an intuitive status of stable, warning, and severe, and corresponding actionable operation and maintenance actions are output (such as increasing the frequency of inspections in the warning state and draining fluid immediately in the severe state). No secondary interpretation is required, and it can be directly connected to operation and maintenance execution.

[0054] Support for optimized resource allocation: By analyzing the status distribution of all gas wells (e.g., how many wells are in a critical state and how many are in an early warning state), gas fields can prioritize the allocation of personnel and equipment to high-risk wells, thus avoiding resource waste (e.g., in Example 1, well A2 does not require additional inspection, and resources can be concentrated on well A1).

[0055] Dynamic weighting adapts to operational priorities: It supports adjusting the weighting of indicators based on the phased operational goals of the gas field (such as focusing on preventing liquid accumulation in a certain quarter or ensuring production in a certain quarter), ensuring that the scoring direction is consistent with the operational needs and enhancing the support value of technology for actual business.

[0056] In summary, the advantage of this invention is not simply multi-index calculation, but rather that it addresses the full-process needs of shale gas well operation and maintenance by creating a closed-loop support system from data input to operation and maintenance decision-making through scenario-based design, precise calculation, forward-looking prediction, and practical output, effectively solving the core pain points of traditional solutions in shale gas well scenarios.

[0057] (3) Historical memory + cross-well comparison to achieve early risk prediction and individualized assessment.

[0058] By introducing historical memory weights and comparing with baseline data from similar wells, scoring distortion caused by fluctuations in single-well data is avoided, while also reflecting the impact of long-term trends.

[0059] Beneficial effects: Risk scoring can identify the decline trend of production potential 1-2 months in advance, increase the proportion of wells in the early warning state by 30%, and transform operation and maintenance decisions from passive response to proactive prevention.

[0060] (4) Three-stage status output + visual suggestions directly drive operation and maintenance decisions.

[0061] The system uses a dual-indicator mapping to a three-stage status, provides anomaly details and suggestions, and supports system integration.

[0062] Beneficial effects: The decision-making efficiency of the operation and maintenance team is improved by 70%, the health of a single well recovers by an average of 15 points after adjustment, and unnecessary operations are reduced by 20% annually.

[0063] (5) The expected benefits and commercial value of the technical solution of this invention after transformation are as follows: After deployment, it can improve operation and maintenance response efficiency by 65%, early warning accuracy by over 90%, reduce annual production losses by 35 million cubic meters, and generate economic benefits of hundreds of millions of yuan. The system is easily integrated with existing platforms and has high promotional value.

[0064] (6) The technical solution of this invention fills a technical gap in the industry both domestically and internationally: Existing general scoring schemes lack shale gas-specific coupling logic and cross-well correction. This invention proposes for the first time a complete scoring framework that combines historical memory, operating condition identification, and cross-well baseline.

[0065] (7) The technical solution of the present invention solves a technical problem that people have long desired to solve but have never been able to solve successfully: Shale gas operation and maintenance scoring has long faced challenges such as operating condition coupling deviation, single-well distortion, and prediction lag. This invention successfully solves these problems through a multi-layer correction mechanism.

[0066] (8) The technical solution of the present invention overcomes technical bias: Breaking through biases such as "health only requires a single indicator" and "cross-well comparison is not applicable to individual wells," this study demonstrates the feasibility of multidimensional dynamic correction. Attached Figure Description

[0067] Figure 1 This is a flowchart of the multi-dimensional health and risk score calculation method for operation and maintenance assessment provided in the embodiments of the present invention.

[0068] Figure 2 This is a schematic diagram of the health-risk dual indicators provided in an embodiment of the present invention.

[0069] Figure 3 This is a structural diagram of the multi-dimensional health and risk scoring calculation system for operation and maintenance assessment provided in this embodiment of the invention.

[0070] Figure 4 This is a comparison of annual production loss (reduction of 35 million cubic meters) provided in the embodiments of the present invention.

[0071] Figure 5 This is the well A2 misjudgment correction process provided in the embodiment of the present invention (eliminating short-term operation interference).

[0072] Figure 6 This invention provides a production trend deviation analysis for well B3 (achieving proactive prediction).

[0073] Figure 7 This is a three-segment state distribution diagram of gas well health and risk dual indicators provided in an embodiment of the present invention.

[0074] Figure 8 This is the health recovery effect after single-well operation and maintenance intervention provided in the embodiments of the present invention.

[0075] Figure 9 This is a comparison of the number of unnecessary tasks per month provided by the embodiments of the present invention (an average annual reduction of approximately 32.6%). Detailed Implementation

[0076] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0077] This invention is a multi-dimensional health and risk scoring calculation method and system designed for shale gas well operation and maintenance scenarios. The core of the method is to output a dual health and risk index and a three-stage production status of stable, early warning, and severe through multi-dimensional index aggregation and operating condition coupling analysis. This provides data support for shale gas well operation and maintenance decisions and resource allocation. Unlike general equipment health scoring schemes, this method focuses on reflecting the coupling characteristics of shale gas well operating conditions (such as the abnormal correlation between liquid accumulation and pressure difference, and the short-term impact of operational events on risk).

[0078] The input data covers all scenarios of shale gas well operation and maintenance, providing a basis for scoring calculations, and specifically includes the following four categories: Real-time operating parameters: real-time gas well production and process parameters (oil pressure, casing pressure, valve pressure, etc.).

[0079] Historical operation and maintenance data: historical output data, historical alarm records, and historical operation event records.

[0080] Abnormal monitoring data: frequency of abnormal alarms, abnormal data on pressure difference, and probability data of fluid accumulation.

[0081] Decision support data: basic information on similar wells (parameter baselines of wells with the same technology and production cycle), and key weighting of gas field operation and maintenance.

[0082] like Figure 1 As shown, during the operation and maintenance assessment of shale gas wells, the sources of various raw signals are diverse, with different dimensions and significant differences in time scale. If directly used for scoring calculations, they are prone to introducing noise interference and occasional anomalies, leading to distortion of health and risk scores. Therefore, this method performs unified and systematic data processing on the signal data before scoring calculations to ensure the stability and interpretability of the subsequent assessment results.

[0083] The system unifies and aligns the collected real-time operating parameters, historical maintenance data, anomaly monitoring data, and auxiliary decision-making data with time. For high-frequency continuous signals such as real-time output, pressure, and differential pressure, a fixed time window is used for sliding sampling, and missing values ​​are filled by linear interpolation or neighborhood mean. For discrete event-type data such as historical alarms and work events, the system maps the event occurrence time to the corresponding statistical period and converts it into frequency, duration, or influence weight forms to achieve a unified time-series representation of heterogeneous signals.

[0084] Outlier suppression and smoothing are performed on signals that have completed time alignment. For indicators such as production and pressure fluctuations that are susceptible to instantaneous disturbances, median filtering or exponentially weighted moving average methods are introduced to reduce the interference of short-term noise on the overall trend judgment. For discrete indicators such as abnormal alarm frequency and pressure difference anomalies, reasonable upper limit cutoff rules are set to prevent extreme outliers from having a nonlinear amplification effect on the scoring results.

[0085] Building upon this foundation, the process moves to the standardization and grading stage. Based on shale gas well operation and maintenance experience and commonly used industry thresholds, bin boundaries are set for indicators such as liquid accumulation probability, valve post-valve pressure fluctuation, pressure difference anomaly degree, alarm recovery time, and impact of operational events. Continuous signals are mapped to standardized score intervals with finite levels, thereby eliminating dimensional differences and improving the comparability between different indicators. This grading result preserves the trend of indicator changes while avoiding the instability of continuous values ​​in extreme ranges on the scoring model.

[0086] A historical memory weighting mechanism is introduced to correct the index scores over time. For each index, when calculating the score for the current period, the index performance from multiple historical periods is superimposed, and decreasing weights are assigned according to time distance. This gives recent data a higher weight in the score, while the influence of long-term data gradually diminishes. This reflects the cumulative effect of the long-term operating status of shale gas wells on the current health and risk level, and avoids excessive interference from single-period anomalies on the overall assessment results.

[0087] After completing the single-indicator processing, the sub-dimensional scoring calculation stage begins. Based on the three core operational dimensions of health, risk, and production potential, the scores of the corresponding indicators are weighted and aggregated according to preset weights to form intermediate scores for each sub-dimensional. Subsequently, according to the gas field's operational priorities, the scores for each sub-dimensional are dynamically weighted and aggregated to obtain initial health and risk scores.

[0088] Before outputting the final score, the score is further corrected by combining the cross-well comparison baseline and the operating condition pattern identification results. By comparing the current well with similar wells of the same process and production cycle, the deviation caused by the special operating conditions of a single well is corrected; at the same time, according to the identified operating condition type, the risk score is applied with corresponding correction factors to make the score results more consistent with the real operation and maintenance scenario.

[0089] After multi-stage signal processing and scoring correction, a stable and interpretable dual health and risk index is output and mapped to the corresponding production and operation status, providing a reliable basis for shale gas well operation and maintenance decisions.

[0090] The calculation method of this invention consists of four core steps: index preprocessing, multi-dimensional aggregation, score correction, and status output, which fully incorporate the characteristics of shale gas well operating conditions and the influence of historical data.

[0091] 1. Indicator preprocessing: laying the foundation for score calculation.

[0092] This step standardizes the input multi-dimensional metrics to ensure the accuracy of subsequent calculations, and it mainly involves two key operations: Indicator bucket boundary setting: Based on shale gas well operation and maintenance experience and industry standards, bucket boundaries are set for indicators such as liquid accumulation probability, valve post-valve pressure fluctuation, pressure difference abnormality, alarm recovery time, and operation events (e.g., liquid accumulation probability of 0%-20% is a low-risk bucket, 20%-50% is a medium-risk bucket, etc.), transforming continuous indicators into quantifiable graded data.

[0093] Historical memory weighting is introduced: each indicator is assigned a historical memory weight. When calculating the current indicator score, the influence of indicator data within the historical period (such as the last 3 months or the last 6 months) is superimposed. The weight decays over time (the weight of recent data is higher), reflecting the cumulative effect of long-term operation and maintenance data on the current score.

[0094] 2. Multi-dimensional aggregation: Calculate health and risk indicators.

[0095] Based on the preprocessed indicators, a weighted scoring model is used to aggregate them into dual core indicators of health and risk, which is done in two steps: Sub-dimensional scoring calculation: Weighted scoring algorithms are designed for the three core operation and maintenance dimensions: health, liquid accumulation risk, and production potential. For example, the health dimension is based on the deviation of production trend (by calculating the deviation between the actual monthly production curve and the Arps decreasing model; the smaller the deviation, the higher the score), and scores are added for indicators such as valve downstream pressure fluctuation and alarm recovery time; the risk dimension uses liquid accumulation probability, pressure difference anomaly, and abnormal alarm frequency as core indicators, and scores are added for the impact of operational events.

[0096] Dual-indicator aggregation: Based on the gas field's operational priorities, the weights of sub-dimensions such as health, liquid accumulation risk, and production potential are dynamically adjusted (e.g., if a gas field is currently focusing on drainage, the weight of liquid accumulation risk is increased). The final health score (0-100 points, the higher the score, the better the health status) and risk score (0-100 points, the higher the score, the higher the risk level) are obtained through weighted summation.

[0097] 3. Scoring Correction: Aligned with the actual operating conditions of shale gas wells.

[0098] By employing two correction methods—cross-well comparison and operational condition identification—data bias from a single well is eliminated, ensuring that the scoring accurately reflects the actual operation and maintenance scenarios of shale gas wells.

[0099] Cross-well comparison baseline correction: Establish a comparison baseline for similar wells with the same technology and production cycle. Compare the current well's health and risk indicators with the baseline data (such as the average health score and average risk score of similar wells). If the deviation exceeds the preset threshold (such as ±15%), the current score is corrected according to the baseline ratio to avoid score distortion caused by special working conditions of a single well.

[0100] Operating condition pattern recognition and correction: Through rule engine or machine learning model, the current operating condition of gas well is automatically identified (such as severe liquid accumulation and well shutdown in the early stage of stable production). The operating condition result is used as a correction factor for risk score (such as risk score increase by 20% under severe liquid accumulation condition, and risk score decrease by 10% under stable production condition), reflecting the direct impact of operating condition coupling characteristics on risk.

[0101] 4. Status output: Converted into results that can be directly used by operations and maintenance personnel.

[0102] like Figure 2 As shown, based on the revised health-risk dual indicators, a three-stage production state is mapped, with the specific rules as follows: Stable status: Health score ≥ 80 points, risk score ≤ 30 points, no serious abnormal indicators, no need for emergency operation and maintenance intervention.

[0103] Warning status: Health score of 60-80 points, or risk score of 30-60 points, with 1-2 medium-risk indicators present, requiring close monitoring and preparation of operation and maintenance plans.

[0104] Critical condition: Health score < 60 points, or risk score > 60 points, indicating critical abnormalities such as severe fluid accumulation or high pressure difference, requiring immediate operation and maintenance (such as drainage or process adjustment).

[0105] like Figure 3 As shown, the system, as the implementation platform of the method, adopts a three-layer architecture design to ensure the efficient operation of the computation process.

[0106] Data layer: Responsible for the collection, storage and cleaning of input data, supporting real-time data access (such as real-time parameter transmission from IoT devices) and historical data archiving, and connecting to similar baseline databases.

[0107] Computation layer: Deploys core computing modules, including an index bucketing engine, a historical weight calculation engine, a weighted scoring model, a cross-well comparison engine, and a working condition identification model, to achieve fully automated calculation of the preprocessing-aggregation-correction process.

[0108] Output layer: Provides a visual output interface to display the values ​​of health and risk indicators, three-stage production status, details of abnormal indicators and operation and maintenance suggestions. It supports data export (such as Excel format) and integration with the operation and maintenance system (such as automatically pushing early warnings / critical statuses to the operation and maintenance work order system).

[0109] Decision support: The output health and risk scores serve as the core basis for determining whether gas wells require intervention (drainage, process adjustment), avoiding blind operation and maintenance.

[0110] Resource optimization: By allocating operation and maintenance resources rationally through the three-stage status distribution of all gas wells (such as prioritizing the allocation of personnel / equipment to wells in critical condition).

[0111] Trend prediction: Based on the analysis of historical memory weight and production trend deviation, the decline in gas well health or the rise in risk can be predicted in advance, realizing the transformation from passive treatment to proactive prevention.

[0112] Combination Figure 4 — Figure 9 It can be seen that this invention is not a simple superposition of existing gas well monitoring or operation and maintenance methods, but rather a new intelligent operation and maintenance solution with significant technical effects at the production decision-making level through multi-indicator fusion modeling, dynamic correction and proactive prediction mechanism, demonstrating outstanding creativity.

[0113] like Figure 4 As shown, under the same production conditions, the annual production loss using the traditional method is approximately 55 million cubic meters, while the annual production loss is significantly reduced to approximately 20 million cubic meters after applying the method of this invention, resulting in a cumulative reduction of approximately 35 million cubic meters. This effect does not stem from the optimization of a single parameter, but rather from the fact that this invention achieves a fundamental improvement in systemic problems such as "false shutdowns, false repairs, and delayed handling" through continuous assessment of the gas well's operating status and optimization of intervention strategies.

[0114] Figure 5 Taking well A2 as an example, this invention demonstrates its technical advantages in correcting misjudgments. Traditional static threshold early warning systems are easily affected by short-term operational disturbances, leading to inflated risk scores. This invention introduces multi-dimensional state aggregation and time continuity constraints to automatically identify and correct short-term operational disturbances, allowing the risk assessment results to return to the actual working conditions, thus eliminating the inherent defect of "alarms upon operation" from a fundamental mechanism perspective.

[0115] Figure 6 This further demonstrates the proactive predictive capability of the present invention. For well B3, the present invention does not trigger a response only after a significant drop in production; instead, it identifies potential anomalies in advance by analyzing the evolution of the deviation between the predicted curve and the actual production trend, thus shifting from "post-event response" to "pre-event warning." This trend deviation-based analysis method breaks through the traditional technical approach that relies on single-point data or experience-based judgment.

[0116] Figure 7 A three-stage state distribution model based on a dual-indicator system of gas well health and risk was constructed, mapping complex operating conditions into three state intervals: stable, warning, and severe, thus enabling clear differentiation of gas wells with different risk levels. This dual-indicator coupled discrimination mechanism upgrades operation and maintenance decisions from fuzzy experience-based judgments to quantifiable and tiered technical solutions.

[0117] Figure 8The results show that after implementing targeted operation and maintenance interventions under the guidance of this invention, the health of multiple gas wells has significantly improved, with an average increase of about 15 points. This proves that the method can not only identify problems, but also effectively guide operation and maintenance actions and form a positive feedback loop.

[0118] at last, Figure 9 The comprehensive benefits of this invention were verified at the operational level. By reducing misjudgments and ineffective interventions, the number of unnecessary operations per month decreased significantly, with an average annual reduction of approximately 32.6%. This not only reduced maintenance costs but also prevented secondary disturbances to the gas well status caused by human operations.

[0119] In summary, this invention, through the synergistic effects of multi-indicator fusion, dynamic correction, trend deviation analysis, and state classification, has formed a new technical system in the field of gas well production monitoring and operation and maintenance decision-making. It solves the long-standing problems of high misjudgment rate, delayed response, and crude intervention in existing technologies, and has outstanding substantive features and significant technological progress.

[0120] Example 1: Based on the coupling characteristics of operating conditions and cross-well comparison, avoid misjudgment of the risk of liquid accumulation in shale gas wells.

[0121] Background: Block A of a shale gas field has 20 gas wells with the same production process (horizontal well fracturing) and the same production cycle (1.5 years after commissioning). Wells A1 and A2 both exhibited abnormal data in [Month] 202X, with a post-valve pressure fluctuation of 0.8 MPa (exceeding the general equipment alarm threshold of 0.5 MPa). Using a traditional general equipment health assessment scheme, both wells were classified as being in a warning state based solely on the single indicator of post-valve pressure fluctuation, and manual inspections at the same frequency (twice a week) were planned, without distinguishing between actual risk differences.

[0122] Application Process: Index Aggregation and Operating Condition Coupling Analysis: This invention incorporates multi-dimensional data from two wells—Well A1, in addition to pressure fluctuations, also exhibits a 38% probability of fluid accumulation (classified as medium risk at the bucket boundary) and has had no operational events in the past month; Well A2, on the other hand, shows no abnormal probability of fluid accumulation and has just completed wellhead valve maintenance work in the past week (the short-term impact of operational events on risk). Through multi-index aggregation, the preliminary risk score for Well A1 is calculated to be 52 points, and the preliminary risk score for Well A2 is 41 points.

[0123] Historical memory weighting correction: Introducing historical data from the past 3 months, well A1 has experienced 2 pressure difference anomalies in the past 3 months (after the historical memory weighting is added, the risk score is corrected to 57 points); well A2 has no abnormal records in the past 3 months (the historical memory weighting offsets part of the impact of short-term operations, and the risk score is corrected to 38 points).

[0124] Cross-well baseline correction: The post-valve pressure fluctuation + fluid accumulation probability of 20 similar wells in Block A were compared with the baseline (average risk score of similar wells is 35 points). After correction, the risk score of well A1 was 55 points (significantly higher than the baseline), and the risk score of well A2 was 36 points (close to the baseline).

[0125] Status output and operation and maintenance decisions: Well A1 is finally determined to be in an early warning state (the frequency of inspections needs to be increased to 4 times a week, and a liquid accumulation monitoring instrument needs to be deployed), and Well A2 is determined to be in a stable state (maintaining a routine inspection once a week is sufficient).

[0126] Example 2: Proactive prevention of declining health of shale gas wells based on historical memory and trend prediction.

[0127] Background: Well B3 in Block B of a shale gas field has been in production for two years with no historical anomalies. However, starting in the Xth quarter of 202X, maintenance personnel discovered that its monthly production had decreased from 80 × 10 4 m³ slowly decreased to 72 × 10 4 m³, the traditional approach only considers output not falling below 60×10 4 The emergency threshold of m³ was determined to be in a stable state, and no intervention measures were taken.

[0128] Application Process: Production Trend Deviation Calculation: This invention uses the Arps declining model (a commonly used production forecasting model in the industry) to calculate the deviation between the actual production curve and the predicted curve of well B3—predicting that the monthly production of this well in quarter X should stabilize at 78 × 10 4 Approximately m³, actual output 72×10⁻⁶ m³ 4 m³, deviation of 7.69% (exceeding the 5% reasonable deviation threshold set for the health-output dimension), initially reducing the output dimension score in the health rating from 90 to 72.

[0129] Multi-indicator aggregation and historical memory correction: Overlaying other dimensions of indicators - the alarm recovery time of well B3 in the past two months has increased from 1 hour to 3 hours, and the frequency of pressure difference anomalies has increased from 0 times per month to 1 time. Combined with the historical data of the past two years (the well's production was stable in the early stage of production, and its health has shown a slow downward trend in the past six months, after the historical memory is superimposed), the health score has been corrected from the initial 75 points to 68 points, and the risk score has been corrected from 25 points to 32 points.

[0130] Operating condition pattern identification and correction: The machine learning model identifies the current operating condition of well B3 as the initial stage of production potential decline (not a sudden anomaly). This operating condition is used as a risk score correction factor, and the risk score is further corrected to 35 points (close to the warning state threshold of 30 points).

[0131] Status Output and Proactive Maintenance: Well B3 was ultimately determined to be in a stable-early warning transition state. Based on this result, the maintenance team adjusted the wellhead gas injection parameters in advance (instead of waiting for production to fall below the threshold). One month later, the monthly production of well B3 rebounded to 76×10 4 m³, health score rebounded to 78 points.

[0132] The multidimensional health and risk scoring calculation system for operation and maintenance assessment of the present invention serves as the engineering implementation of the method. It adopts a three-layer architecture of data layer, calculation layer and output layer to realize automated analysis and auxiliary decision support for gas well operation status.

[0133] The data layer is responsible for the acquisition, storage, and preprocessing of raw data. Data sources include real-time operating parameters collected from gas well sites and historical production data. It supports real-time parameter access via IoT devices and archives historical data in a baseline database of similar wells. Before the data enters the computation layer, the data layer performs integrity checks, outlier removal, time alignment, and format standardization on the acquired data to ensure the availability and consistency of the input data, thus providing a standardized data foundation for subsequent calculations.

[0134] The computing layer deploys core computing modules to automate the calculation of health and risk scores. First, the indicator bucketing engine segments and categorizes indicators based on their historical distribution and preset rules. Then, the historical weight calculation engine calculates the weight coefficients of each indicator at different time scales, combining historical trends and memory decay mechanisms. Based on this, the weighted scoring model performs a comprehensive weighted calculation of each indicator, generating the health and risk scores for a single well. Simultaneously, the cross-well comparison engine compares the current gas well with baselines of similar wells to identify the degree of deviation. The operating condition identification model further identifies and classifies the current operating condition of the gas well by combining the scoring results with key indicator patterns. Through the coordinated operation of these modules, the entire process from preprocessing and aggregation calculation to deviation correction is automated.

[0135] The output layer is used to visualize the calculation results and interface with the system. The output content includes the values ​​of health and risk indicators, the corresponding three-stage production status, the details of abnormal indicators, and operation and maintenance suggestions. It also supports exporting the results as a spreadsheet file or pushing them to the operation and maintenance work order system via an interface to realize the automatic distribution of early warning information.

[0136] In practical applications, the system's output health and risk scores serve as the basis for decision-making regarding whether to implement maintenance operations such as well drainage and process adjustments, thereby avoiding blind actions based on experience. Simultaneously, the system prioritizes maintenance resources through the distribution of all well statuses and anticipates declining health and rising risk trends based on historical weights and production trend deviations, shifting from reactive measures to proactive prevention.

[0137] This invention integrates the method into programs and system modules in computer devices, computer-readable storage media, and information data processing terminals, enabling consistent collaborative operation among the method, system, device, and storage media. This allows for the calculation and application of multi-dimensional health and risk scores for operation and maintenance assessment.

[0138] It should be noted that embodiments of the present invention can be implemented in hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by a suitable instruction execution system, such as a microprocessor or dedicated-design hardware. Those skilled in the art will understand that the above-described devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuitry such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field-programmable gate arrays, programmable logic devices, etc., or by software executed by various types of processors, or by a combination of the above-described hardware circuitry and software, such as firmware.

[0139] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, and within the spirit and principles of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A multi-dimensional signal processing method for shale gas well operation and maintenance assessment, characterized in that, Includes the following steps: The acquired multi-dimensional operation and maintenance signals are time-aligned so that continuous and discrete signals at different time scales are uniformly mapped to the same statistical period. Anomaly suppression processing is performed on maintenance signals that are aligned with completion time to reduce the impact of transient disturbances on signal trends; Based on the preset index grading boundaries, the processed continuous signal is converted into a graded score signal; When calculating the graded score for the current period, historical memory weights are introduced to weight and sum the scores of similar signals within historical periods. The further away the historical period is from the current period, the smaller its weight.

2. The method according to claim 1, characterized in that, The time alignment process includes: Continuous signals are sampled and statistically analyzed according to a fixed time window, and discrete event signals are converted into frequency or duration within the time window.

3. The method according to claim 1, characterized in that, The anomaly suppression process includes performing sliding smoothing on continuous signals and setting upper limit truncation rules for discrete signals.

4. A method for calculating health and risk scores for shale gas well operation and maintenance assessment, characterized in that, Includes the following steps: Based on the processed multi-dimensional operation and maintenance signals, a health score sub-model and a risk score sub-model are constructed respectively. The health score sub-model uses the production change trend as the core variable, and superimposes scores related to pressure fluctuations and alarm recovery. The risk scoring sub-model uses fluid accumulation-related indicators and pressure difference abnormality-related indicators as core variables, and superimposes the impact score of abnormal events; The outputs of the health score sub-model and the risk score sub-model are weighted and aggregated to obtain the health score and the risk score, respectively.

5. The method according to claim 4, characterized in that, In the weighted aggregation, the weights of each scoring sub-model are set according to the key configurations of gas field operation and maintenance.

6. The method according to claim 4, characterized in that, The output change trend is obtained by calculating the degree of deviation between the current cycle output and the historical benchmark output.

7. A scoring correction method for shale gas well operation and maintenance assessment, characterized in that, Includes the following steps: Establish a baseline for scoring similar gas wells under the same technological conditions; The health score and risk score of the target gas well are compared with the baseline score. When the score deviates from the scoring baseline by more than a preset threshold, the score of the target gas well is corrected according to the baseline ratio; The current operating mode is identified based on the gas well operation signals, and an operating mode correction factor is applied to the risk score based on the identification results.

8. The method according to claim 7, characterized in that, The term "same type of gas well" refers to a collection of gas wells with the same production cycle and process parameters.

9. The method according to claim 7, characterized in that, The operating mode is determined through rule-based judgment or model recognition.

10. The method according to claim 7, characterized in that, The revised health score and risk score are mapped to a stable state, a warning state, or a severe state.

Images