Cluster well group efficient group mining method

Abstract

The invention relates to a cluster well group efficient group mining method. The method comprises the steps that a stratum structure, heat reservoir distribution, ground temperature gradient, permeability and porosity of a target area are obtained; generating well group collaborative logic based on geological parameters and underground sensor data, wherein the well group collaborative logic comprises dynamic well spacing calculation, well group Voronoi clustering and grouping, a yield distribution relation and an interference suppression strategy; production optimization parameters are determined according to collaborative logic, temperature, pressure and flow data are collected in real time through a monitoring system, and production is dynamically adjusted in combination with a geological interference model (including a heat-flow coupling equation). Layered well spacing and selective perforation are implemented for multiple heat reservoirs, and parameters of each layer are independently set. Monitoring data interference is filtered through collaborative logic, when a single well fluctuates, the yield and the recharge amount of an adjacent well are automatically adjusted to keep the total capacity stable, a fault diagnosis model is updated based on real-time data, parameters are recursively optimized, and closed-loop control is formed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of geothermal cluster irrigation, in particular to a high-efficiency cluster extraction method for cluster wells. BACKGROUND

[0002] Geothermal resources, as a clean and renewable energy source, have broad development prospects. In geothermal development, cluster wells are a common drilling layout method, which involves concentrating multiple wells in a region to achieve large-scale exploitation of geothermal resources. However, current cluster well extraction methods have some problems, resulting in low extraction efficiency. For example, the cooperative work between wells in the cluster is not effective, and the extraction strategy cannot be dynamically adjusted according to the actual situation of the stratum, resulting in low utilization of geothermal resources. At the same time, during the extraction process, the monitoring and control of the cluster are not precise enough, which can easily cause inter-well interference, rapid energy decay of the thermal reservoir, and other problems, affecting the long-term stable operation of the cluster well group. Therefore, there is an urgent need for a method that can improve the extraction efficiency and stability of cluster wells. SUMMARY

[0003] The present application aims to provide a high-efficiency cluster extraction method for cluster wells to solve the problems of low extraction efficiency and poor stability of existing cluster wells mentioned in the background. The present application aims to provide a high-efficiency cluster irrigation method for cluster wells, including the following steps:

[0004] A high-efficiency cluster extraction method for cluster wells, including the following steps:

[0005] Obtain the geological parameters of the target area, including stratum structure, thermal reservoir distribution, geothermal gradient, permeability, and porosity;

[0006] Determine the production optimization parameters based on the geological parameters and well cluster coordination logic; the well cluster coordination logic includes well production allocation relationship, inter-well interference suppression strategy, and coordination control rules;

[0007] According to the production optimization parameters, real-time acquisition of downhole temperature, pressure, flow rate, and thermal reservoir state data is performed through a monitoring and control system to dynamically adjust the production of the well cluster.

[0008] Further, the generation of the well cluster coordination logic also depends on downhole sensor data, including real-time acquisition values of temperature, pressure, and flow rate sensors.

[0009] Further, the well cluster coordination logic includes:

[0010] Adjust the inter-well spacing based on permeability and porosity: increase the well spacing when the permeability is high, and decrease the well spacing when the permeability is low;

[0011] Cluster the well group according to the dynamic well spacing to form a coordinated extraction unit to minimize inter-well interference.

[0012] Furthermore, it also includes: excluding invalid well location schemes based on geological parameters and well group coordination logic; the invalid well locations include well group layouts that do not cover thermal reservoirs or whose surface conditions do not meet drilling requirements.

[0013] Furthermore, the determination of the production optimization parameters includes: constructing a geological interference model by combining sensor data and geological parameters; the model includes the influence of geothermal gradient direction and heat flow path on inter-well interference.

[0014] Furthermore, when constructing the geological interference model, interference noise data caused by thermal reservoir blockage or well leakage anomalies are filtered out.

[0015] Furthermore, for multi-layered geothermal reservoirs, layered well placement is implemented based on a geological disturbance model, and selective perforation technology is used to precisely exploit the target strata; production optimization parameters are set independently according to the characteristics of each layered geothermal reservoir.

[0016] Furthermore, the geothermal gradient and permeability distribution map are integrated into the geological disturbance model to improve the spatial representation accuracy of real-time monitoring data.

[0017] Furthermore, based on well group collaboration logic, interference filtering is implemented on the monitoring data: duplicate well group interference alarms are suppressed, and low-confidence abnormal data that conform to the collaboration rules are retained.

[0018] Furthermore, it also includes: controlling inter-well collaborative production through production optimization parameters and sensor data: when the production capacity of a single well fluctuates, automatically adjusting the production and reinjection volume of adjacent wells to keep the total production capacity of the well group within a set range;

[0019] The fault diagnosis model is updated based on real-time data to identify equipment failures or thermal reservoir blockages.

[0020] Dynamic iterative optimization: The updated production parameters are recursively applied to the allocation of reinjection volume and production adjustment to form a closed-loop control.

[0021] Beneficial effects include:

[0022] A precise mapping from geological attributes to engineering parameters is achieved through well spacing calculation formulas. For high-permeability zones (permeability > 100 mD), well spacing is increased to 500-800 meters (e.g., 600 meters for sandstone layers with a permeability of 120 mD), reducing the number of wells by 50% and avoiding overlapping production capacity. For low-permeability zones (permeability < 10 mD), well spacing is reduced to 200-300 meters (e.g., 250 meters for carbonate rock layers with a permeability of 5 mD), increasing the number of wells to compensate for the production capacity of individual wells, thereby increasing the total production capacity by 40%. A collaborative production unit is formed by combining the Voronoi spatial partitioning algorithm with well clustering based on permeability similarity (difference < 20%). Within the unit, a "main well-auxiliary well" pressure linkage mechanism is implemented (e.g., auxiliary wells reduce production by 10% simultaneously when the main well pressure drops by 0.3 MPa), and reinjection wells are deployed at the unit boundary to isolate interference.

[0023] A thermal-fluid coupled mathematical model (including fluid continuity equations and heat conduction equations) was constructed, and reservoir dynamics were simulated using a finite element method (FEM) solver. This guided the implementation of differentiated well placement strategies: in shallow, high-permeability, high-temperature zones (depth < 2000 meters, temperature > 150℃), a cluster of horizontal wells with a depth of 500-1000 meters was used; in deep, low-permeability zones (depth > 3000 meters), directional wells with a 30°-60° inclination angle were deployed, and staged fracturing was implemented. Intelligent completion tools were applied to achieve selective perforation, and production rates were adjusted according to permeability ratios using an in-hole flow control device (ICD) (e.g., a production ratio of 10:5:3 for 100mD:50mD:30mD layers). Production parameters were set independently for each layer: a pressure drop of 1-2 MPa was allowed in high-permeability layers, and cold water was injected into low-temperature reinjection layers (20-30℃) to enhance energy replenishment. Parameters were dynamically optimized every 12 hours.

[0024] A distributed fiber optic temperature measurement system, piezoelectric pressure gauges, and ultrasonic flow meters are deployed to form a high-frequency monitoring network. The IEEE 1588 protocol is used to achieve multi-well data synchronization (latency <1 millisecond). A static geological model (including geothermal gradient field and permeability tensor matrix) is fused with real-time data, and the well network is optimized using permeability anisotropy. Kalman filtering is used to dynamically correct model biases. The control layer applies a model predictive control algorithm to adjust equipment parameters 24 hours in advance, forming a closed loop of "monitoring-diagnosis-control". Wavelet transform (db4 wavelet 3-level decomposition) is used to filter out high-frequency noise >0.1Hz, accurately identifying reservoir blockage characteristics.

[0025] A triple-coordination mechanism is established: ① Dynamic production allocation model (the production of a 150mD well is three times that of a 50mD well); ② Staggered production strategy (well groups are operated in cyclical units, such as group 1 producing for 4 hours and then group 2 taking over), extending the reservoir pressure recovery period by 50%; ③ Inter-well linkage response (when the pressure of well A drops by 10%, the adjacent well B automatically reduces production by 15% and the injection well increases injection by 20%). The alarm system merges redundant alarms (such as merging the pressure drop alarms of 3 wells within a unit into 1) through spatiotemporal correlation rules (10-minute time window + 500-meter spatial radius), while retaining low-confidence abnormal data that conforms to the heat flow path prediction model (sensor confidence <60% still triggers diagnosis).

[0026] A dual-effect judgment model for geological and engineering aspects is established: Geological ineffectiveness includes uncovered thermal reservoirs (thickness < 20 meters) or fault zones (permeability < 5 mD and continuity < 50%); engineering ineffectiveness includes ground slope > 15°, distance from fault < 100 meters, or exceeding the drilling rig's capability depth. The judgment logic is embedded in the genetic algorithm's fitness function: F = 0.6 × productivity score + 0.3 × effectiveness score - 0.1 × invalid well penalty (0.5 points deducted for each invalid well). For example, a scheme containing 2 invalid wells is eliminated because the penalty value = 1. Attached Figure Description

[0027] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0028] 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.

[0029] Reference Figure 1 To achieve the above objectives, this application adopts a group mining method based on "geological parameter analysis - collaborative logic generation - dynamic control" to eliminate control systems that rely solely on fixed well networks or human experience.

[0030] Specifically, a cluster well group efficient production method includes the following steps: acquiring geological parameters of the target area, including formation structure, thermal reservoir distribution, geothermal gradient, permeability, and porosity; determining production optimization parameters based on geological parameters and well group collaboration logic; the well group collaboration logic includes the production distribution relationship of each well, inter-well interference suppression strategy, and collaborative control rules; and dynamically adjusting well group production by collecting downhole temperature, pressure, flow rate, and thermal reservoir status data in real time through a monitoring and control system according to the production optimization parameters.

[0031] In the above embodiments, geological parameter acquisition is a fundamental step in the mass extraction method, and its technical characteristics are reflected in multi-dimensional data acquisition and correlation analysis of geological attributes. Specifically, this includes stratigraphic structure (such as the distribution boundaries between sedimentary layers and bedrock, and stratigraphic dip angles), spatial morphology of thermal reservoirs (thickness, depth, and continuity), geothermal gradient (the rate of temperature change per kilometer of depth, directly affecting the value of thermal energy development), permeability (characterizing the ability of fluids to pass through rock, measured in millidarcy mD), and porosity (the proportion of pore volume in rock, affecting the thermal and fluid storage capacity of thermal reservoirs). For example, in thermal reservoir exploration, a three-dimensional geothermal gradient model can be constructed by combining seismic exploration and well logging data. If the geothermal gradient is 3℃ / km, then the temperature at a depth of 2000m is 60℃ higher than the surface, providing a basis for well depth design.

[0032] In the above embodiments, permeability and porosity directly determine the well network layout strategy. In high-permeability zones (k>100mD), where fluid flow is strong, the well spacing needs to be increased to 500-800m to avoid the "competition effect" of fluid between multiple wells; in low-permeability zones (k<10mD), the well spacing needs to be reduced to 200-300m, compensating for insufficient production capacity of a single well by increasing the number of wells. The geothermal gradient guides the priority of thermal energy extraction, with production wells being deployed preferentially in high-temperature areas (>150℃).

[0033] In the above embodiments, the well group coordination logic is the core of achieving efficient ensemble production, and its technical features include three dimensions: ① Dynamic production allocation: Differentiated production capacity configuration is implemented based on the heterogeneity of the thermal reservoir. By establishing a permeability-production mapping model, high-permeability wells can be allocated higher production, while low-permeability wells have their production limited to avoid excessive pressure drop. For example, in a well group, well A has a permeability of 150 mD, and well B has a permeability of 50 mD, then the production of well A is set to be 3 times that of well B, balancing the overall production capacity. ② Multi-method coordination for interference suppression: Pressure field superposition caused by simultaneous extraction from multiple wells is avoided through staggered extraction. For example, the well group operates in a cyclical mode of "group 1 extracts for 4 hours → group 2 takes over extraction", which extends the thermal reservoir pressure recovery cycle by 50%. Reservoir pressure is maintained by injecting fluid through reinjection wells. When the pressure drop of the production well group exceeds a set threshold, the reinjection wells are automatically activated, and the pressure decay is suppressed by dynamically adjusting the injection volume. ③ Intelligent collaborative control rules: Define an inter-well linkage response mechanism. For example, when the pressure of well A drops by 10%, the adjacent well B will automatically reduce production by 15%, while the injection well will increase the injection volume by 20%. The reservoir pressure balance is maintained through the dual regulation of "production-injection".

[0034] In the above embodiments, dynamic closed-loop control achieves a fully automated process from data acquisition to command execution. This is mainly reflected in the following aspects: a real-time monitoring system is designed, collecting data through downhole temperature sensors, pressure sensors, and electromagnetic flowmeters, and transmitting it wirelessly to the surface control system. An integrated control algorithm is employed: a model predictive control algorithm is used to predict the pressure and temperature change trends over the next 24 hours based on a geological disturbance model, adjusting valve opening and pump speed in advance. A data-model fusion mechanism is designed to match and fuse static geological parameters with real-time monitoring data, correcting model errors using a Kalman filter algorithm. For example, if the initial geological model predicts a permeability of 80 mD for a certain area, but real-time flow data shows lower-than-expected production capacity, the system automatically corrects the permeability of that area to 60 mD, improving subsequent control accuracy.

[0035] In some embodiments, the generation of the well group coordination logic also relies on downhole sensor data, including real-time acquisition values ​​from temperature, pressure, and flow sensors.

[0036] In the above embodiments, the core of sensor data reliance lies in constructing a high-frequency, multi-source data acquisition network. For example, a distributed fiber optic temperature measurement system is used for temperature sensors to monitor the temperature distribution throughout the entire wellbore in real time and identify geothermal reservoir sections and crossflow channels. Piezoelectric pressure gauges are deployed downhole to capture transient changes in the pressure field (such as pressure fluctuations caused by inter-well interference). Ultrasonic flow meters are used, installed in the wellhead pipeline, with a measurement accuracy of ±0.5%, to monitor geothermal fluid production and reinjection in real time. All sensor data is transmitted to the surface via downhole cables or wireless acoustic waves, using the IEEE 1588 clock synchronization protocol to ensure that the timestamp error of multi-well data is <1ms, providing a timing reference for inter-well interference analysis. For example, when the flow rate of well A drops sharply, it can be determined through synchronized data whether it is caused by increased production from well B.

[0037] In some embodiments, the well group collaboration logic includes: dynamically adjusting the well spacing based on permeability and porosity: increasing the well spacing when permeability is high and decreasing the well spacing when permeability is low; and clustering the well group according to the dynamic well spacing to form collaborative mining units to minimize inter-well interference.

[0038] In the above embodiments, the well spacing calculation formula is D = K·φ / k, where D is the well spacing (m), k is the permeability (mD), φ is the porosity (%), and K is an adjustment coefficient (ranging from 0.5 to 2, determined by the reservoir temperature and fluid viscosity). Higher permeability k indicates stronger fluid diffusion, requiring a larger well spacing to avoid interference; higher porosity φ indicates stronger thermal storage capacity, allowing for a more efficient reduction in well spacing to fully utilize energy storage. For example, when the reservoir temperature k = 150mD, φ = 10%, and K = 0.8, the well spacing D = 0.8 × 150 / 10 = 12m. In actual engineering, this needs to be adjusted to the 200-300m range based on surface conditions.

[0039] In some embodiments, the well spacing classification strategy includes: High permeability zones (k>100mD): Well spacing is increased to 500-800m. For example, in a sandstone thermal reservoir with k=120mD, the well spacing is set at 600m, reducing the number of wells by 50% compared to the traditional 300m well spacing, while avoiding overlapping production capacity. Low permeability zones (k<10mD): Well spacing is reduced to 200-300m, compensating for single-well production capacity by increasing the number of wells. For example, in a carbonate thermal reservoir with k=5mD, the well spacing is 250m, the well group density is increased by 2 times compared to the conventional approach, and the total production capacity is increased by 40%.

[0040] In some embodiments, clustering well groups based on dynamic well spacing to minimize interference is achieved primarily through the following method: Using well locations as base points, Voronoi polygons are generated, with fluid flowing to the corresponding well within each polygon, ensuring that the "influence domains" between wells do not overlap. For example, after dividing a group of 10 wells using Voronoi, the control area of ​​each well is clearly defined, avoiding fluid competition. Based on the distance between wells and the similarity of permeability, the well group is divided into several clusters. A neighborhood radius of 1.5 times the well spacing is set, and the minimum sample size is 3 wells. Wells that are spatially adjacent and have a permeability difference of <20% are grouped into the same unit.

[0041] In some embodiments, wells within the same unit share a pressure control target, employing a "main well + auxiliary well" model. The main well adjusts its production based on pressure changes, and the auxiliary well automatically follows suit. For example, when the pressure in the main well within the unit decreases by 0.3 MPa, the auxiliary well simultaneously reduces its production by 10% to maintain pressure balance within the unit. Low-production wells or reinjection wells are deployed at the unit boundaries as buffer zones to isolate the pressure fields of different units.

[0042] In some embodiments, traditional methods, which rely on experience to set well spacing, cannot adapt to the heterogeneity of thermal reservoirs. This method achieves automatic mapping of "geological attributes - well spacing" through quantitative formulas. Clustering and grouping, which combines spatial topology and geological parameters, better reflects fluid flow patterns than simple geometric grouping (such as rectangular grids) and effectively suppresses inter-well interference.

[0043] In some embodiments, invalid well location schemes are excluded based on geological parameters and well group coordination logic. Invalid well locations include well group layouts that do not cover the thermal reservoir or whose surface conditions do not meet drilling requirements. Geological invalidity determination includes missing thermal reservoir cover and insufficient thermal reservoir continuity. Specifically, missing thermal reservoir cover: 3D geological modeling is used to verify whether the well trajectory penetrates the top and bottom plates of the thermal reservoir. For example, if a well's designed trajectory only passes through the overlying strata of the thermal reservoir and does not enter the target reservoir (thickness ≥ 20m), it is determined to be geologically invalid. Insufficient thermal reservoir continuity: seismic attribute analysis (such as coherence volume, amplitude anomalies) is used to identify thermal reservoir fault zones. If the well location is located in a fault zone causing reservoir fragmentation (permeability < 5mD and continuity < 50%), it is determined to be invalid. Project ineffectiveness is determined by ground condition limitations and insufficient drilling feasibility. Ground condition limitations include: a ground slope >15°, which significantly increases well site construction costs and drilling risks; wellbore collapse is likely to occur near fault zones (<100m from the fault); and drilling is directly excluded from residential areas, nature reserves, and other prohibited drilling areas. Insufficient drilling feasibility is defined as: when the expected drilling depth exceeds the equipment capacity (e.g., a designed well depth of 5000m, but the maximum drilling depth of the rig is 4000m), or when the formation leakage risk level is >3 (leakage >50m³). 3 If the value is / h), the project is deemed invalid. The invalidation logic is embedded into optimization algorithms such as Genetic Algorithm (GA) to achieve automatic filtering. For example, the following formula can be used: F = w1·F 产能 +w2·F 有效性 -w3·P 无效 F represents the projected production capacity score of the well group (weight w1 = 0.6); F 有效性 The effectiveness of the well site is scored (1 point for geological and engineering compliance, 0 points otherwise, weight w2 = 0.3); P 无效 This is an invalidity penalty (0.5 points are deducted for each invalid well, with a weight w3 = 0.1). For example, if a well group scheme contains 2 invalid wells, then P_invalid = 2 × 0.5 = 1, which significantly reduces the fitness and causes the scheme to be eliminated in the iteration.

[0044] In some embodiments, determining the production optimization parameters includes: constructing a geological interference model by combining sensor data and geological parameters; the model includes the influence of geothermal gradient direction and heat flow path on inter-well interference. When constructing the geological interference model, interference noise data caused by thermal reservoir blockage or well leakage anomalies is filtered out.

[0045] Geothermal gradient direction: A heat flow path vector field is generated using geothermal logging data to guide the matching of well cluster layout with the heat flow direction. For example, when the heat flow direction is NE-SW, the optimal well cluster arrangement is NE-SW to reduce heat flow obstruction and improve thermal energy extraction efficiency by 15%. Permeability tensor matrix: Considering the anisotropy of the thermal reservoir, a permeability tensor is constructed, as shown in the following formula:

[0046]

[0047] Where kx, ky, and kz represent the permeability in the x, y, and z directions, respectively. For example, if the measured values ​​for a thermal reservoir are kx = 80 mD, ky = 50 mD, and kz = 20 mD, then the fluid flow capacity is strongest in the x direction, and the well spacing in the x direction can be increased by 20%.

[0048] The thermal-fluid coupling equations are solved using the finite element method (FEM), and the governing equations are as follows:

[0049] Fluid continuity equation:

[0050] Heat conduction equation:

[0051] Where P is pressure, T is temperature, S is reservoir compressibility, ρc is heat capacity, and Kc is thermal capacity. T Let be the thermal conductivity and Q be the heat source term. Dynamic simulation of the pressure and temperature fields is achieved using software such as COMSOL.

[0052] Actual monitoring data contains a large amount of interference noise, requiring filtering through multiple mechanisms, including identification of thermal reservoir blockage characteristics: Anomaly in pressure-flow combination: Under normal interference, pressure and flow fluctuate periodically (e.g., pressure drops and slowly recovers after inter-well interference), while blockage manifests as a continuous rise in pressure (>0.5 MPa / day) accompanied by a sudden drop in flow (>20%). Multi-sensor verification of well leakage faults: Data conflict detection: If the surface flow sensor shows a 10% decrease in flow, but the wellbore level gauge shows a synchronous drop in level (normally, a decrease in flow corresponds to an increase in level), this is identified as well leakage. This data set is used for noise filtering and triggers the well leakage processing flow. Wavelet transform denoising algorithm: The pressure and flow data are decomposed into three levels using db4 wavelets to remove high-frequency noise (frequency >0.1 Hz) and retain trend signals. For example, well pressure data may contain high-frequency vibration noise (frequency 0.5 Hz) caused by drilling operations; after wavelet transform, the trend of pressure changes (frequency 0.01 Hz) is preserved.

[0053] In some embodiments, for multi-layered thermal reservoirs, well placement is implemented in layers based on a geological disturbance model, and selective perforation technology is used to precisely exploit the target layers; production optimization parameters are set independently according to the characteristics of the layered thermal reservoirs.

[0054] In the above embodiments, multi-layered thermal reservoir stratification control resolves inter-layer conflicts during synergistic production. Its technical characteristics are reflected in differentiated well deployment: First, the thermal reservoir depth is matched with the well type: High-permeability shallow layers (<2000m): Cluster horizontal wells are used, with horizontal sections 500-1000m long, parallel to the thermal reservoir strike, increasing the contact area with the reservoir. Low-permeability deep layers (>3000m): Directional wells are deployed, with inclination angles of 30°-60°, penetrating multiple thermal reservoirs, and permeability is increased through staged fracturing. Thermal reservoir properties and exploitation priority: Shallow high-temperature zones (>150℃) are prioritized for exploitation, utilizing high-temperature fluids for direct power generation; deep low-temperature zones (<100℃) serve as reinjection target layers to maintain reservoir energy balance. For example, in multi-layered thermal reservoirs, production wells are deployed in the shallow 180℃ area, and reinjection wells are deployed in the deep 80℃ area.

[0055] In the above embodiments, selective perforation enables directional extraction from the target formation. Essentially, mature technologies employ intelligent completion tools to achieve segmented perforation, with each perforation segment equipped with a flow control device to automatically adjust the production rate based on the formation permeability. For example, if a well penetrates three thermal reservoirs with permeabilities of 100 mD, 50 mD, and 30 mD respectively, the ICD device ensures a production rate ratio of 10:5:3 for the three segments, preventing the high-permeability layer from "rushing" to extract fluids from the low-permeability layer.

[0056] Production parameters are independently optimized for each geothermal reservoir layer, primarily including differentiated pressure control. For high-permeability layers (k>100mD), a pressure drop of 1-2 MPa is allowed, maintained through reinjection; for low-permeability layers (k<30mD), the pressure drop is limited to ≤0.5 MPa to prevent further decrease in permeability. The reinjection strategy is layered: the high-temperature production layer receives reinjection at a temperature ≥60℃ to prevent reservoir cooling; the low-temperature reinjection layer is injected with cold water (20-30℃) to increase reservoir fluid density and enhance energy replenishment. Dynamic iterative optimization: an independent production model is established for each layer, and parameters are updated every 12 hours based on real-time data.

[0057] In some embodiments, geothermal gradient and permeability distribution maps are integrated into the geological disturbance model to improve the spatial representation accuracy of real-time monitoring data. The geothermal gradient contour maps and permeability distribution maps are rasterized and converted into input grids for the numerical model. For example, the geothermal gradient of the thermal reservoir increases from 2.5℃ / km in southwest to northeast to 3.8℃ / km, while the permeability decreases from 100mD in west to east to 30mD. This is rasterized into a 50m×50m grid, with each grid containing geothermal and permeability parameters. A temperature field coupling term is added to the traditional pressure-flow model, and a thermal convection term is added to the governing equations. Where u is the fluid velocity, it realizes the bidirectional coupling of temperature and fluid flow, and more realistically reflects the dynamics of the thermal reservoir.

[0058] Based on inter-well sensor data (such as temperature and pressure), a thermal map of the reservoir state is generated using the Kriging interpolation algorithm. For example, temperature data from 10 wells, after Kriging interpolation, generates a temperature distribution cloud map of the entire reservoir, improving the spatial resolution from well spacing (200-600m) to 50m, visually displaying high-temperature and low-temperature anomaly zones. Real-time monitoring data is compared with model predictions hourly; if the error exceeds 10%, model parameters are automatically adjusted. For instance, if the model predicts a temperature of 95℃ in a certain area, but the actual measured temperature is 88℃, the system corrects the thermal conductivity of that area from 2.5W / (m·K) to 2.8W / (m·K), improving the accuracy of subsequent predictions.

[0059] The system uses thermal mapping to identify "cold spots" (temperature drop > 5°C) and "thermal breakthroughs" (temperature rise > 10°C) in hot reservoirs in real time, guiding well group adjustments. For example, when a cold spot is detected in a certain area, the system automatically shuts down nearby production wells and starts heating in the reinjection wells.

[0060] In some embodiments, interference filtering is implemented on monitoring data based on well group collaboration logic: duplicate well group interference alarms are suppressed, and low-confidence abnormal data that conforms to the collaboration rules are retained. When multiple wells in the same collaboration unit trigger alarms concurrently, the cause is attributed to a single geological event. For example, if three wells in the unit trigger "pressure drop" alarms simultaneously, it is determined to be a regional pressure fluctuation (rather than an independent fault), and only one alarm record is generated to avoid interference from duplicate information for maintenance personnel. Time-space correlation determination: An alarm time window (e.g., 10 minutes) and a spatial threshold (e.g., a radius of 500m) are set, and similar alarms within the same time window and spatial threshold are merged. If, within 5 minutes, well A (x=100, y=200) and well B (x=150, y=220) successively trigger "flow anomaly" alarms, and the distance between the two wells is 50m < 500m, the alarms are merged into a "regional flow anomaly" alarm, and the location is determined to be the area between the two wells.

[0061] The above embodiments further include: controlling inter-well collaborative production through production optimization parameters and sensor data: when the production capacity of a single well fluctuates, automatically adjusting the production and reinjection volume of adjacent wells to keep the total production capacity of the well group within a set range; updating the fault diagnosis model based on real-time data to identify equipment failures or thermal reservoir blockage; and dynamic iterative optimization: recursively applying the updated production parameters to the allocation of reinjection volume and production adjustment to form a closed-loop control. In the above embodiments, collaborative production control suppresses production capacity fluctuations. Specifically, when the production deviation of a single well ΔQ > ±15%, a PID controller is used to adjust the production of adjacent wells, and the reinjection volume is adjusted synchronously according to the reinjection coefficient β = 1.2 to ensure that the total production capacity fluctuation of the well group is < ±5%.

[0062] Production parameters are updated every 6 hours. The objective function is to minimize the sum of the deviations between actual and predicted production. The constraint is that the reservoir pressure drop does not exceed the critical value (≤2MPa for high-permeability layers and ≤0.5MPa for low-permeability layers). Closed-loop process: Real-time data acquisition → Fault diagnosis model update → Anomaly confirmation → Adjustment of adjacent well production and reinjection volume → Model predictive control (MPC) pre-adjustment of future state → Parameter recursive optimization → Enter the next cycle.

[0063] In the foregoing description, examples have been described with reference to specific exemplary embodiments. However, it will be apparent that various modifications and changes can be made to the specific examples without departing from the scope set forth in the appended claims, and the claims are not limited to the specific examples described above.

Claims

1. A method for efficient mass production of cluster wells, characterized in that, Includes the following steps: Obtain geological parameters of the target area, including stratigraphic structure, distribution of thermal reservoirs, geothermal gradient, permeability, and porosity; Based on geological parameters and well group collaboration logic, production optimization parameters are determined; the well group collaboration logic includes the production distribution relationship of each well, inter-well interference suppression strategy, and collaborative control rules. Based on production optimization parameters, the monitoring and control system collects downhole temperature, pressure, flow rate, and thermal reservoir status data in real time, and dynamically adjusts well group production.

2. The efficient cluster well production method according to claim 1, characterized in that, The generation of the well group coordination logic also relies on downhole sensor data, including real-time acquisition values ​​from temperature, pressure, and flow sensors.

3. The efficient cluster well production method according to claim 2, characterized in that, The well group coordination logic includes: The well spacing is dynamically adjusted based on permeability and porosity: the well spacing is increased when the permeability is high and decreased when the permeability is low. Well groups are clustered and grouped according to dynamic well spacing to form collaborative mining units and minimize inter-well interference.

4. The efficient cluster well production method according to claim 1, characterized in that, Also includes: Based on geological parameters and well group coordination logic, invalid well location schemes are excluded; the invalid well locations include well group layouts that do not cover thermal reservoirs or whose surface conditions do not meet drilling requirements.

5. The efficient cluster well production method according to claim 1, characterized in that, The determination of the production optimization parameters includes: A geological disturbance model is constructed by combining sensor data and geological parameters; the model includes the influence of geothermal gradient direction and heat flow path on inter-well disturbance.

6. The efficient cluster well production method according to claim 5, characterized in that, When constructing a geological disturbance model, filter out noise data caused by thermal reservoir blockage or well leakage anomalies.

7. The efficient cluster well production method according to claim 6, characterized in that, For multi-layered geothermal reservoirs, well placement is implemented in layers based on a geological disturbance model, and selective perforation technology is used to precisely exploit the target layers; production optimization parameters are set independently according to the characteristics of each layer of geothermal reservoir.

8. The efficient cluster well production method according to claim 1, characterized in that, Integrating geothermal gradient and permeability distribution maps into a geological disturbance model improves the spatial representation accuracy of real-time monitoring data.

9. The efficient cluster well production method according to claim 1, characterized in that, Based on well group collaboration logic, interference filtering is implemented on monitoring data: duplicate well group interference alarms are suppressed, and low-confidence abnormal data that conforms to the collaboration rules are retained.

10. The efficient cluster well production method according to claim 1, characterized in that, Also includes: By using production optimization parameters and sensor data, the system controls coordinated production between wells: when the production capacity of a single well fluctuates, it automatically adjusts the production and reinjection volume of adjacent wells to keep the total production capacity of the well group within a set range. The fault diagnosis model is updated based on real-time data to identify equipment failures or thermal reservoir blockages. Dynamic iterative optimization: The updated production parameters are recursively applied to the allocation of reinjection volume and production adjustment to form a closed-loop control.

Images