Method and system for monitoring oil well production based on tracer injection concentration

Abstract

The present application relates to the technical field of oil well channeling monitoring based on tracer, and particularly relates to an oil well production monitoring method and system based on tracer injection concentration. The present application firstly records a total response sequence of pulse concentration; further compares the injection history record of background tracer with the output concentration, and constructs a dynamic hydrodynamic baseline parameter set; further singular value decomposes the current total response sequence of pulse concentration, and reconstructs a single-path concentration response sequence; further constructs a hydrodynamic characteristic set according to each single-path concentration response sequence and the dynamic hydrodynamic baseline parameter set; further conducts a pairwise Granger causality test on all single-path concentration response sequences, and constructs a concurrent channeling path system characteristic set in combination with the hydrodynamic characteristic set; and finally generates an output water source composition report, quantitatively reveals the output water source, accurately reflects the transient channel structure of the oil reservoir, and provides a reliable basis for water plugging management and injection-production optimization.

Description

Technical Field

[0001] This invention relates to the field of tracer-based monitoring of oil well crossflow technology, specifically to a method and system for monitoring oil well production based on tracer injection concentration. Background Technology

[0002] The inherent heterogeneity of the reservoir medium, especially the presence of advantageous channels such as natural or artificial fractures and high-permeability bands, makes it easy for the injected polymer solution to flow along these advantageous channels, causing cross-flow. Without sufficient oil displacement, it quickly and inefficiently circulates to the production well, which not only wastes chemical agents but also seriously reduces the economic benefits of the overall oil displacement project.

[0003] Current methods generally rely on infrequent, periodic sampling of production wells to indirectly assess development effectiveness by analyzing macroscopic indicators such as polymer concentration in the produced fluid. These methods essentially involve delayed observations of the reservoir's average response over a long timescale. They cannot distinguish between concurrent crossflow systems activated by a single injection event and composed of multiple paths with distinct characteristics, nor can they determine the topological relationships between these paths. Consequently, the diagnosis of crossflows and the assessment of production remain at a general and vague level, leading to costly and inefficient remediation solutions. Summary of the Invention

[0004] To address the problem that existing monitoring and analysis of crossflow events lacks depth, leading to high costs and low efficiency in remediation solutions, the present invention aims to provide an oil well production monitoring method and system based on tracer injection concentration. The specific technical solution adopted is as follows:

[0005] A method for monitoring oil well production based on tracer injection concentration, the method comprising:

[0006] Continuous and constant injection of background tracer and injection of pulse tracer based on injection line pressure fluctuations; monitoring the production concentrations of background tracer and pulse tracer;

[0007] Based on a preset noise threshold, the output concentration of the pulse tracer is compared, and the total pulse concentration response sequence is recorded. Within the current preset historical neighborhood, the injection history of the background tracer is compared with the output concentration to construct a dynamic hydraulic baseline parameter set. Singular value decomposition is performed on the current total pulse concentration response sequence to reconstruct the single-path concentration response sequence. Based on each single-path concentration response sequence and the dynamic hydraulic baseline parameter set, a hydraulic feature set is constructed. A pairwise Granger causality test is performed on all single-path concentration response sequences, and combined with the hydraulic feature set, a concurrent crossflow path system feature set is constructed.

[0008] Within the current preset monitoring period, the total liquid production and total water production are monitored. By combining all the pulse concentration total response sequences and the corresponding concurrent crossflow path system feature sets, the ineffective circulation volume of pore-type and fracture-type in the total ineffective circulation volume is analyzed, and a report on the composition of the produced water source is generated.

[0009] Furthermore, the method for obtaining the dynamic hydraulic baseline parameter set includes:

[0010] The two sequences of the injection history of the background tracer and the output concentration are deconvolved to obtain the average transit time and the time variance of the response curve of the centroid of the background tracer in the preset historical neighborhood. The results are then input into the exponential moving average filter to output the dynamic baseline transit time and dynamic baseline dispersion, which constitute the dynamic hydraulic baseline parameter set.

[0011] Furthermore, the method for obtaining the single-path concentration response sequence includes:

[0012] The total pulse concentration response sequence is transformed into a hydrodynamic trajectory matrix using time delay embedding technology.

[0013] The hydrodynamic trajectory matrix is ​​subjected to singular value decomposition. The singular values ​​in the resulting diagonal matrix are arranged in descending order. The singular values ​​are accumulated from largest to smallest until the preset energy ratio is reached for the first time. The singular values ​​involved in the accumulation are retained and the number of concurrent paths is determined.

[0014] Based on each retained singular value, its corresponding left singular vector, and right singular vector, a combined reconstruction is performed to obtain the single-path concentration response sequence corresponding to each retained singular value.

[0015] Furthermore, the method for obtaining the hydraulic feature set includes:

[0016] For each of the single-path concentration response sequences: the relative velocity index is obtained based on the difference between the time from injection to the peak concentration and the average transit time; the relative diffusion index is obtained based on the difference between the time variance of the single-path concentration response sequence and the dynamic baseline diffusion; the proportion of the square of the singular value to the sum of the squares of the singular values ​​of all the single-path concentration response sequences is taken as the path dominance rate.

[0017] The relative velocity index, the relative dispersion index, and the path dominance rate constitute the hydraulic feature set of the single-path concentration response sequence.

[0018] Furthermore, the method for obtaining the feature set of the concurrent streaming path system includes:

[0019] Perform pairwise Granger causality tests on all the single-path concentration response sequences to obtain a causal relationship matrix; use the sum of all element values ​​corresponding to the predicted relationship in the causal relationship matrix for each single-path concentration response sequence as the convergence center score; normalize the sum of all off-diagonal elements in the causal relationship matrix and use the normalized result as the system convergence index.

[0020] The convergence center score, the system convergence index, the total number of single-path concentration response sequences, and the hydraulic feature set of each single-path concentration response sequence are used to construct a concurrent crossflow path system feature set.

[0021] Furthermore, the method for generating the produced water source composition report includes:

[0022] Within the current preset monitoring period, each pulse injection event is treated as a target event.

[0023] For the target event, the total pulse concentration response sequence and the synchronously acquired total production volume are fused to obtain the total mass of the pulse tracer for the target event; the ratio of the total recovered mass of the pulse tracer to the total mass of the pulse tracer at the time of injection is multiplied by the total mass of all tracers injected during the pulse tracer injection process to obtain the total volume of the ineffective cycle; the total volume of the ineffective cycle is divided into each single-path concentration response sequence according to the path dominance rate to obtain the single-path leakage volume;

[0024] When the system convergence index is less than or equal to the preset convergence index threshold, the single-path leakage volume is added to the pore-type ineffective circulation volume and the crack-type ineffective circulation volume based on the relative dispersion index.

[0025] When the system convergence index is greater than the preset convergence index threshold, the single-path leakage volume corresponding to the largest relative dispersion index is taken as the main channel invalid circulation volume, and the single-path leakage volume corresponding to the remaining relative dispersion indices is accumulated and added to the upstream tributary invalid circulation volume.

[0026] A report on the composition of produced water sources is generated based on the total water production within a preset monitoring period and the various invalid cycles of all the target events.

[0027] Furthermore, the method of adding the single-path leakage volume to the pore-type ineffective circulation volume and the crack-type ineffective circulation volume based on the relative dispersion index includes:

[0028] The relative dispersion index is classified based on a preset dispersion threshold. The sum of the single-path leakage volumes corresponding to the relative dispersion indices that are greater than or equal to the preset dispersion threshold is taken as the porosity-type invalid circulation volume. The sum of the single-path leakage volumes corresponding to the relative dispersion indices that are less than the preset dispersion threshold is taken as the crack-type invalid circulation volume.

[0029] Furthermore, the method for recording the total pulse concentration response sequence includes:

[0030] When the output concentration of the pulse tracer first exceeds a preset noise threshold, the output concentration sequence of the pulse tracer is recorded until the output concentration is lower than the preset noise threshold and remains stable for a preset duration, at which point the recording stops, and the total pulse concentration response sequence is obtained.

[0031] Furthermore, the method for injecting pulsed tracers based on injection pipeline pressure fluctuations includes:

[0032] When the injection pipeline pressure reaches a preset pressure ratio of the average pressure of the preset first neighborhood, the pulse tracer is injected, and the injection continues for a preset single injection duration before stopping.

[0033] The present invention also proposes an oil well production monitoring system based on tracer injection concentration. The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any of the steps of the oil well production monitoring method based on tracer injection concentration described above.

[0034] The present invention has the following beneficial effects:

[0035] This invention first employs a dual tracer strategy to separate data sources, facilitating more accurate identification of abnormal events. It then records the total pulse concentration response sequence based on a preset noise threshold, avoiding misinterpretation of small fluctuations caused by measurement jitter or instrument drift as channel responses, thus narrowing the analysis scope. Furthermore, it constructs a dynamic hydraulic baseline parameter set to provide a reference standard for subsequent analysis. Next, it performs singular value decomposition on the current total pulse concentration response sequence to reconstruct single-path concentration response sequences, capturing a more detailed channel structure. Finally, it constructs a hydraulic feature set based on each single-path concentration response sequence in conjunction with the dynamic hydraulic baseline parameter set, automatically eliminating systematic drift caused by overall reservoir evolution and improving the resolution and interpretability of crossflow diagnosis. Furthermore, it performs pairwise Granger causality tests on all single-path concentration response sequences, combining them with the hydraulic feature set to construct a concurrent crossflow path system feature set, providing a complete and quantitative input basis for subsequent analysis of the produced water source composition. Finally, it integrates the concurrent crossflow path system feature set and monitoring data to generate a produced water source composition report. This invention, through combined injection and high-frequency monitoring, utilizes noise threshold screening, dynamic baseline, singular value decomposition, and Granger causality test to identify concurrent crossflow paths and extract features, quantitatively reveal the source of produced water, accurately reflect the transient channel structure of the reservoir, and provide a reliable basis for water shut-off management and injection-production optimization. Attached Figure Description

[0036] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 A flowchart illustrating an oil well production monitoring method based on tracer injection concentration, provided as an embodiment of the present invention;

[0038] Figure 2 A flowchart illustrating a method for obtaining a single-path concentration response sequence according to an embodiment of the present invention;

[0039] Figure 3 This is a flowchart illustrating a method for generating a report on the composition of produced water sources, as provided in one embodiment of the present invention. Detailed Implementation

[0040] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of an oil well production monitoring method and system based on tracer injection concentration proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0041] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0042] The following description, in conjunction with the accompanying drawings, details the specific scheme of an oil well production monitoring method and system based on tracer injection concentration provided by the present invention.

[0043] Please see Figure 1 The diagram illustrates a flowchart of an oil well production monitoring method based on tracer injection concentration according to an embodiment of the present invention, specifically including:

[0044] Step S1: Continuously inject background tracer and inject pulse tracer according to the pressure fluctuation of the injection pipeline; monitor the production concentration of background tracer and pulse tracer.

[0045] Because the overall pressure and saturation fields of an oil reservoir evolve slowly during long-term water injection development, the main migration characteristics of the inter-well fluid are not constant. Using an early, static parameter as a criterion will inevitably lead to misjudgments of later events. Furthermore, the opening or penetration of dominant channels caused by injection pressure fluctuations is instantaneous. Therefore, in one embodiment of this invention, a dual tracer injection strategy is employed to separate the "background" from the "event" at the data source.

[0046] On the surface injection pipeline of the injection well, a dual tracer injection unit controlled by a central system is deployed. This injection unit is responsible for executing two independent injection logics: continuously injecting a background tracer and injecting a pulsed tracer according to pressure fluctuations in the injection pipeline.

[0047] An online monitoring unit is installed at the wellhead of the production well corresponding to the injection well to synchronously record and timestamp data at a high acquisition frequency of minutes. Because the dominant channels opened in the formation due to pressure fluctuations may be transient, their response duration in the production well may be very short. Conventional low-frequency sampling would completely miss or distort key morphological information of such events, and subsequent signal separation algorithms require high-resolution data to construct an effective analysis matrix. The monitoring unit needs to collect the following three time-series data: the production concentration C_A,prod(t) of background tracer A, the production concentration C_B,prod(t) of pulse tracer B, and the total production volume Q_total(t) of the production well.

[0048] Preferably, in one embodiment of the present invention, in order to continuously track the changes in the macroscopic seepage characteristics of the reservoir body, the system continuously injects an inert background tracer A into the injection fluid at a known, long-term constant mass flow rate, thereby maintaining a tracer background field with a gradually changing concentration that can be tracked throughout the entire area affected by the well.

[0049] To specifically identify fluids in dominant channels that may be activated due to abnormally high pressure conditions, the system links the injection logic of pulse tracer B with the injection line pressure P_inj(t). When the system detects that the pressure of P_inj(t) is high within a set time window, it automatically triggers an injection of pulse tracer B.

[0050] Specifically: When the injection pipeline pressure reaches a preset pressure ratio of the average pressure of the preset first neighborhood, the pulse tracer is injected, and the injection continues for a preset single injection duration before stopping.

[0051] As an example, the first neighborhood is preset to the most recent 24 hours, the preset pressure ratio is 110%, and the preset single injection duration is 10 minutes. When the injection pipeline pressure reaches 110% of the average pressure of the most recent 24 hours, there may be a high-pressure condition where formation micro-fractures are opened or dominant channels are eroded. The injection continues for the preset single injection duration and then stops. The system records the injection start time t_inj,i and the total mass M_B_inj,i of the injected pulse tracer for each pulse injection event i. When the injection reaches the preset single injection duration and the conditions for triggering pulse tracer injection are still met, the injection continues for the preset single injection duration.

[0052] It should be noted that, in one embodiment of the present invention, the acquisition frequency of the monitoring unit is set to once every 10 seconds, the background tracer A is NaBr, injected at a constant concentration of 1 mg / L, and the pulse tracer B is fluorobenzoate (3-FBA), injected continuously at a concentration of 30 mg / L for 10 minutes.

[0053] In other embodiments of the present invention, the implementer may adjust the preset first neighborhood, preset pressure ratio, preset single injection duration, types and injection schemes of tracers A and B, and sampling frequency, which will not be elaborated further.

[0054] Step S2: Based on a preset noise threshold, compare the output concentration of the pulse tracer and record the total pulse concentration response sequence; within the current preset historical neighborhood, compare the injection history and output concentration of the background tracer to construct a dynamic hydraulic baseline parameter set; perform singular value decomposition on the current total pulse concentration response sequence to reconstruct the single-path concentration response sequence; construct a hydraulic feature set based on each single-path concentration response sequence in conjunction with the dynamic hydraulic baseline parameter set; perform pairwise Granger causality tests on all single-path concentration response sequences, and combine the hydraulic feature set to construct a concurrent crossflow path system feature set.

[0055] The overall purpose of step S2 is to conduct a progressive analysis, from decomposition, qualitative analysis, quantitative analysis to structural diagnosis, on the complex response signal labeled by pulse tracer B in S1, which may be caused by multiple concurrent dominant channels, in order to reveal the complete systemic characteristics behind the transient spurious event.

[0056] Therefore, the output concentration of the pulse tracer is first compared with the preset noise threshold, and the total pulse concentration response sequence is recorded to avoid false small fluctuations caused by measurement jitter or instrument drift being misjudged as channel responses, thus narrowing the analysis range and providing a basis for subsequent analysis.

[0057] Preferably, in one embodiment of the present invention, during a stable production period without pulse tracer B, the production concentration of B is collected for a period of time (e.g., 24 hours), and based on the 3σ principle, the sum of the mean and three times the standard deviation of this data is used as a preset noise threshold.

[0058] When the output concentration of the pulse tracer first exceeds the preset noise threshold, the output concentration sequence of the pulse tracer is recorded until the output concentration is lower than the preset noise threshold and remains stable for a preset duration, at which point the recording stops and the total pulse concentration response sequence is obtained.

[0059] The preset stabilization time is 5 minutes. In other embodiments of the present invention, a preset noise threshold can also be set based on empirical values, and the preset stabilization time can be adjusted, which will not be elaborated further.

[0060] When the total response sequence of pulse concentration is recorded, it indicates the presence of an abnormal high-pressure operating event, and in-depth analysis is required.

[0061] Since the macroscopic seepage characteristics of the reservoir will slowly drift with the development process, it is necessary to establish a reference benchmark that can be updated adaptively over time in order to accurately isolate the "abnormal" attributes of transient events. Therefore, within the current preset historical neighborhood, the injection history and production concentration of the background tracer are compared to construct a dynamic hydraulic baseline parameter set to provide a reference standard for the future. Here, the current corresponds to the occurrence time of the current event (recording the total response sequence of the pulse concentration).

[0062] Preferably, in one embodiment of the present invention, the preset historical neighborhood is the historically adjacent 24 hours; by performing deconvolution calculation on the two sequences of the injection history record of the background tracer and the production concentration, the "impulse response function (response curve)" of the reservoir to the injection signal can be recovered. The average transit time can be obtained through the centroid position of the impulse response function (the weighted average position on the time axis), which represents the average migration time of the fluid through the conventional pore network of the reservoir body; the dispersion can be obtained through the time variance of the impulse response function, which represents the degree of response broadening (the degree of dispersion and diffusion) caused by convection and dispersion when the fluid passes through the conventional pore network of the reservoir body.

[0063] Furthermore, considering that the results of a single deconvolution are susceptible to noise, short-term anomalies, or local disturbances, the exponential moving average filter can weight and smooth the transit time and dispersion, allowing the parameters to be updated slowly over time. This preserves the long-term trend while reducing occasional fluctuations, ensuring the continuity and stability of the dynamic baseline.

[0064] Therefore, by using the deconvolution calculation results, the average transit time and time variance of the response curve of the background tracer centroid within the preset historical neighborhood are obtained and input into the exponential moving average filter to output the dynamic baseline transit time and dynamic baseline dispersion, which constitute a dynamic hydraulic baseline parameter set, providing a reference for subsequent analysis and making the identification, decomposition and diagnosis of abnormal events more accurate and reliable.

[0065] It should be noted that time variance is the variance of the elements in the time series. Deconvolution and exponential moving average filtering are existing technologies and will not be elaborated further.

[0066] It should be noted that the analysis process for constructing the dynamic hydraulic baseline parameter set is the same for each time. Only one example is described here. The update frequency can be the same as the data acquisition frequency; or a fixed periodic update can be set, and a supplementary update can be performed when the total pulse concentration response sequence is recorded. This will not be elaborated further.

[0067] Because a single high-pressure injection pulse in a heterogeneous formation may simultaneously penetrate or activate a primary fracture and several secondary high-permeability zones, the transport characteristics of these different channels vary greatly. Their tracer responses superimpose at the production wellhead, forming a complex pulse concentration response sequence that cannot be directly analyzed. Therefore, singular value decomposition (SVD) is performed on the current pulse concentration response sequence. Using SVD, the orthogonal main transport components constituting the mixed signal are separated, and the single-path concentration response sequence is reconstructed, capturing a more detailed channel structure.

[0068] Preferably, in one embodiment of the present invention, please refer to Figure 2 The diagram illustrates a flowchart of a method for obtaining a single-path concentration response sequence according to an embodiment of the present invention, specifically including:

[0069] Step S201: The total response sequence of pulse concentration is converted into a hydrodynamic trajectory matrix by time delay embedding technology.

[0070] To apply singular value decomposition, it is necessary to transform the one-dimensional time series into a two-dimensional matrix. Here, the pulse concentration total response sequence is transformed into a hydrodynamic trajectory matrix through time delay embedding technology.

[0071] Specifically, a row vector is constructed by extracting d consecutive data points from the total pulse concentration response sequence and sliding it with a time step τ to generate subsequent rows until the end of the sequence. Here, τ is taken as the value at which the autocorrelation function of the total pulse concentration response sequence first drops to zero, and d is determined by the pseudo-nearest neighbor method to ensure that the dynamic structure can be fully unfolded in the embedding space.

[0072] Step S202: Perform singular value decomposition on the hydrodynamic trajectory matrix, arrange the singular values ​​in the diagonal matrix obtained by decomposition in descending order, accumulate the singular values ​​in descending order until the preset energy ratio is reached for the first time, retain the singular values ​​that participate in the accumulation and determine the number of concurrent paths.

[0073] The preset energy ratio is set to 99%. The magnitude of the singular values ​​reflects the energy proportion of each transport component in the total response signal. By analyzing the energy distribution of the singular value spectrum, those singular values ​​whose cumulative energy contribution reaches 99% of the total energy are retained, and their number k is defined as the number of concurrent paths in this event.

[0074] Specifically, singular values ​​are accumulated from largest to smallest until a preset energy ratio is reached for the first time. The singular values ​​involved in the accumulation are retained and the number of concurrent paths is determined, which quantitatively characterizes the number of concurrent dominant channels activated in this (pulse) injection event.

[0075] Step S203: Based on each retained singular value, the corresponding left singular vector and right singular vector, perform combined reconstruction to obtain the single-path concentration response sequence corresponding to each retained singular value.

[0076] After determining the number of concurrent paths, it is necessary to reduce these abstract mathematical components to response sequences with clear hydraulic significance and perform characteristic calculations on them.

[0077] Specifically, for each single-path concentration response sequence, the outer product matrix of the singular values, the left singular vector, and the transposed right singular vector is then diagonally averaged to obtain the corresponding single-path concentration response sequence. Specifically, during the diagonal averaging, for each time point t of the single-path concentration response sequence, the arithmetic mean of all elements X(p,q) in the outer product matrix that satisfy p + q - 1 = t is taken, where p and q are the horizontal and vertical coordinates in the matrix, respectively.

[0078] Obtaining the single-path concentration response sequence corresponding to each retained singular value can be regarded as the expected concentration change curve (sequence) of the production well assuming that tracer B only passes through a single path. This yields the number of potential independent fluid paths, and the relative importance of a single path is quantified by using singular values.

[0079] It should be noted that time delay embedding technology and singular value decomposition are well-known technologies and will not be elaborated further.

[0080] Each single-path concentration response sequence provides a potential independent fluid transport path, corresponding to a single path, and includes the specific response pattern of that path in the time domain. The dynamic hydraulic baseline parameter set provides a reference system. Based on each single-path concentration response sequence and the dynamic hydraulic baseline parameter set, a hydraulic feature set is constructed to automatically eliminate the systematic drift caused by the overall reservoir evolution and improve the resolution and interpretability of crossflow diagnosis.

[0081] Preferably, in one embodiment of the present invention, for each single-path concentration response sequence: the difference between the time from injection to the appearance of the concentration peak and the average transit time reflects the multiple of the flow rate of the corresponding path relative to the main background (indicator A), thereby obtaining the relative velocity index;

[0082] The difference between the time variance of the single-path concentration response sequence and the dynamic baseline dispersion reflects whether the dispersion / mixing degree of the substance in the channel is abnormal, and the relative dispersion index is obtained.

[0083] Considering that singular values ​​represent the energy proportion (contribution rate) of a path, in order to distinguish the dominance of different single paths, the proportion of the square of the singular value to the sum of the squares of the singular values ​​of all single path concentration response sequences is taken as the path dominance rate.

[0084] The relative velocity index, relative dispersion index, and path dominance rate constitute the hydraulic feature set of the single-path concentration response sequence.

[0085] As an example, the differences between data are represented by a ratio for each single-path concentration response sequence: the average transit time is used as the numerator, the time from injection to the concentration peak is used as the denominator, and the ratio of the fractions is used as the relative velocity index.

[0086] The time variance of the single-path concentration response sequence is used as the numerator, the dynamic baseline dispersion is used as the denominator, and the ratio of the fractions is used as the relative dispersion index. The relative dispersion index is used to distinguish between crack-type flow and pore-type flow. The relative dispersion index of crack-type flow is much less than 1, while the relative dispersion index of pore-type flow is approximately equal to 1.

[0087] To explore whether there are structural connections between these concurrent paths, such as in a complex underground fracture network, multiple upstream secondary channels may eventually converge into a downstream main channel. This "convergence" structure is quite different from multiple unrelated "parallel" channels in terms of subsequent engineering management strategies.

[0088] Therefore, we further paired the single-path concentration response sequences and performed Granger causality tests on all single-path concentration response sequences. Combined with the hydraulic feature set, we constructed a feature set of concurrent crossflow path systems, which provides a complete and quantitative input basis for subsequent analysis of the source composition of produced water.

[0089] Preferably, in one embodiment of the present invention, a pairwise Granger causality test is performed on all single-path concentration response sequences, that is, the single-path concentration response sequences are tested in pairs to obtain a causal relationship matrix; the element M(m,n) in the matrix reflects the degree of contribution of the past value of path m to the predicted future value of path n, where path n corresponds to the prediction relationship. Based on this causal relationship matrix, topological features are further extracted:

[0090] The sum of all element values ​​corresponding to the predicted relationships in the causal relationship matrix for each single-path concentration response sequence is used as the convergence center score. For example, S_hub,i,s represents the convergence center score of the s-th path in the i-th event, where S_hub,i,s = ΣM(m,s), where Σ is the summation symbol, and the summation iterates through all m ≠ s. This score quantifies the probability that the s-th path serves as a downstream "main artery" or "convergence center."

[0091] The sum of all off-diagonal elements in the causal relationship matrix is ​​normalized, and the normalized result is used as the system convergence index. The diagonal elements are those with the same horizontal and vertical coordinates in the causal relationship matrix, representing the causal relationship of a path to itself. The off-diagonal elements are those with different horizontal and vertical coordinates, representing the causal relationship values ​​between the path and other different paths.

[0092] The normalization here is linear normalization, which is constructed by summing all off-diagonal elements in the causal relationship matrix of all historical events into a data dimension, and then normalizing in the data dimension. This is an existing technology and will not be elaborated further.

[0093] Finally, the convergence center score, system convergence index, total number of single-path concentration response sequences, and hydraulic feature set of each single-path concentration response sequence are used to construct a concurrent crossflow path system feature set.

[0094] The feature set of the concurrent crossflow path system includes the number of concurrent paths (total number of single path concentration response sequences), the relative velocity index, relative dispersion index and path dominance rate of each single path, the convergence center score and system convergence index of each path.

[0095] It should be noted that the Granger causality test is a well-known technique and will not be elaborated upon further.

[0096] Step S3: Within the current preset monitoring period, monitor the total liquid production and total water production, and combine the total pulse concentration response sequence and the corresponding concurrent crossflow path system feature set to analyze the ineffective circulation volume of pore-type and fracture-type in the total ineffective circulation volume, and generate a report on the composition of the produced water source.

[0097] In step S2, the total pulse concentration response sequence is decomposed into several single-path concentration response sequences by singular value decomposition and baseline reference, and they are labeled with hydraulic features in the hydraulic feature set, which further provides structural constraints in the feature set of concurrent crossflow path system.

[0098] Ultimately, within the current preset monitoring period, the total produced liquid volume and total produced water volume are monitored. Combining the total response sequence of all pulse concentrations and the corresponding concurrent crossflow path system feature set, and integrating the time domain, structural domain, and volume domain, the ineffective circulation volume of porosity and fracture is analyzed in the total ineffective circulating water volume. A report on the composition of produced water sources is generated, providing path-level quantitative basis for engineering guidance, providing deeper decision-making basis, thereby formulating more suitable treatment solutions, reducing solution costs, and improving solution efficiency.

[0099] Preferably, in one embodiment of the present invention, please refer to Figure 3The diagram illustrates a flowchart of a method for generating a produced water source composition report according to an embodiment of the present invention, specifically including:

[0100] Step S301: Within the current preset monitoring period, each pulse injection event is taken as a target event.

[0101] Each pulse injection event is analyzed. Here, each pulse injection event is taken as the target event. Taking the current (latest) pulse injection event as the target event as an example, the description will not be repeated.

[0102] Step S302: For the target event, fuse the total pulse concentration response sequence and the synchronously acquired total production volume to obtain the total recovered mass of the pulse tracer for the target event; multiply the ratio of the recovered total mass of the pulse tracer to the total mass of the pulse tracer at the time of injection by the total mass of all tracers injected during the pulse tracer injection process, and use this as the total volume of the invalid cycle; divide the total volume of the invalid cycle into each single-path concentration response sequence according to the path dominance rate to obtain the single-path leakage volume.

[0103] To correlate the concurrent crossflow path system feature set obtained in step S2 with the actual product volume, it is necessary to calculate the ineffective circulating water volume for each path. Based on the principle of mass balance, the total ineffective circulating water volume caused by a single complex crossflow event (pulse injection event) is allocated according to the contribution ratio of each path. Specifically:

[0104] First, the total pulse concentration response sequence and the synchronously acquired total liquid production are fused together. The total pulse concentration response sequence C_B,i(t) represents the production concentration sequence of the pulse tracer, and the total liquid production Q_total(t) represents the mass of all liquids. The total recovered mass M_B,rec_total,i of the pulse tracer for the target event (corresponding to index i) is obtained through integration. The calculation formula is as follows: The integration interval extends from the start time to the end time of the total response sequence of the event's pulse concentration.

[0105] The ratio of the total recovered mass of the pulse tracer (M_B,rec_total,i) to the total mass of the pulse tracer at the time of injection (M_B,inj,i) is further multiplied by the total mass of all tracers injected during the pulse tracer injection process (V_inj_pulse,i) to obtain the total volume of the invalid cycle (V_leak_total,i).

[0106] It should be noted that the total pulse concentration response sequence is matched with the most recent historical pulse injection event. In another embodiment of the present invention, the implementer may also obtain the time interval between the start time of the total pulse concentration response sequence and the start time of the historical pulse injection event based on the dynamic baseline transit time, and match the total pulse concentration response sequence corresponding to the time interval closest to the dynamic baseline transit time with the most recent historical pulse injection event. In other embodiments of the present invention, the implementer may use a variety of pulse tracers and use them alternately to reduce the possibility of event matching errors.

[0107] Finally, using the path dominance rate as the contribution ratio, the total volume of invalid cycles is allocated to each single-path concentration response sequence according to the path dominance rate to obtain the single-path leakage volume.

[0108] As an example, the single-path leakage volume is calculated by multiplying the path dominance rate of each single path by the total volume of invalid loops.

[0109] Step S303: When the system convergence index is less than or equal to the preset convergence index threshold, the single-path leakage volume is accumulated into the pore-type ineffective circulation volume and the crack-type ineffective circulation volume based on the relative dispersion index; when the system convergence index is greater than the preset convergence index threshold, the single-path leakage volume corresponding to the largest relative dispersion index is taken as the main channel ineffective circulation volume, and the single-path leakage volume corresponding to the remaining relative dispersion indices is accumulated into the upstream tributary ineffective circulation volume.

[0110] Because the engineering management strategies for "parallel" channel systems and "converging" channel systems are completely different, they must be distinguished at the production monitoring level.

[0111] When the system convergence index is less than or equal to the preset convergence index threshold, it indicates that there is no significant convergence relationship between the paths of the target event, and it is dominated by a parallel channel crossflow system. Based on the relative dispersion index, the single-path leakage volume is added to the pore-type ineffective circulation volume and the crack-type ineffective circulation volume, respectively. Specifically:

[0112] Considering that the relative dispersion index of crack-type flow is much less than 1, and the relative dispersion index of pore-type flow is approximately equal to 1, the relative dispersion index is classified based on a preset dispersion threshold. The sum of the single-path leakage volumes corresponding to the relative dispersion indices greater than or equal to the preset dispersion threshold is taken as the pore-type ineffective circulation volume; the sum of the single-path leakage volumes corresponding to the relative dispersion indices less than the preset dispersion threshold is taken as the crack-type ineffective circulation volume.

[0113] When the system convergence index is greater than the preset convergence index threshold, it indicates that the target event has a downstream trunk (such as multiple cracks merging into a main channel). The single-path leakage volume corresponding to the largest relative dispersion index is taken as the invalid circulation volume of the trunk, and the single-path leakage volume corresponding to the remaining relative dispersion index is accumulated and added to the invalid circulation volume of the upstream tributary.

[0114] It should be noted that, in one embodiment of the present invention, the system convergence index of historical events can be collected, and a distribution histogram can be drawn. In theory, the distribution presents a bimodal distribution, so the system convergence index value corresponding to the valley between the two peaks is taken as the preset convergence index threshold. The relative diffusion index of historical events is collected, and the preset diffusion threshold is obtained by using the maximum inter-class variance method.

[0115] In other embodiments of the present invention, the implementer may set an empirical value as a preset convergence index threshold, and may also establish a numerical model of a typical crossflow system (such as a parallel crack or a tree crack network), simulate and calculate the theoretical value of the system convergence index of each model, and take the point with the highest discrimination as the threshold.

[0116] Step S304: Based on the total water production within the preset monitoring period and the various invalid circulation volumes of all target events, generate a report on the composition of the produced water source.

[0117] Finally, by integrating all the ineffective circulation volumes and combining them with the total produced water volume, the effective displacement water volume can be obtained, and a report on the composition of the produced water source can be generated.

[0118] Specifically, the preset monitoring cycle is set to 24 hours, and a report on the composition of the produced water source is generated every 24 hours, or a report is generated using the data from the most recent 24 hours after each pulse injection event ends.

[0119] The total ineffective circulation volume is the sum of all events accumulated within the preset monitoring period of the current analysis, and the difference between the total production volume and the total ineffective circulation volume is the effective displacement volume.

[0120] The report is broken down into:

[0121] • Effective water displacement: [percentage %]

[0122] • Ineffective circulating water: [Percentage %]

[0123] Pore-related ineffective cycle contribution: [percentage %]

[0124] Contribution to crack-type invalid cycles: [Percentage %]

[0125] Contribution from upstream tributaries: [Percentage %]

[0126] Contribution from main roads: [Percentage %]

[0127] Here, invalid circulating water corresponds to the total invalid circulating water volume. In other embodiments of the present invention, the implementer may set other report formats, which will not be elaborated further.

[0128] The report enables in-depth and dynamic monitoring of the "quality" of oil well production, providing managers with direct and quantitative data support for optimizing injection and production parameters, and providing an evaluation benchmark for assessing the effectiveness of subsequent engineering measures such as profile control and water shut-off, thus forming a complete "monitoring-diagnosis-decision-evaluation" closed loop.

[0129] An embodiment of the present invention also provides an oil well production monitoring system based on tracer injection concentration. The system includes a memory, a processor, and a computer program. The memory is used to store the corresponding computer program, and the processor is used to run the corresponding computer program. When the computer program runs in the processor, it can implement the oil well production monitoring method based on tracer injection concentration described in steps S1-S3.

[0130] In summary, to address the problem that existing monitoring and analysis of crossflow events lack depth, leading to high costs and low efficiency in remediation solutions, this invention provides an oil well production monitoring method and system based on tracer injection concentration. First, this invention compares the production concentration of the pulse tracer with a preset noise threshold and records the total pulse concentration response sequence. Next, it compares the historical injection history of the background tracer with the production concentration to construct a dynamic hydraulic baseline parameter set. Then, it performs singular value decomposition on the current total pulse concentration response sequence to reconstruct single-path concentration response sequences. Finally, it constructs a hydraulic feature set based on each single-path concentration response sequence in conjunction with the dynamic hydraulic baseline parameter set. Next, it performs pairwise Granger causality tests on all single-path concentration response sequences and, combined with the hydraulic feature set, constructs a concurrent crossflow path system feature set. Finally, it integrates the concurrent crossflow path system feature set and monitoring data to generate a report on the composition of produced water sources. This invention, through combined injection and high-frequency monitoring, utilizes noise threshold screening, dynamic baseline, singular value decomposition, and Granger causality test to identify concurrent crossflow paths and extract features, quantitatively reveal the source of produced water, accurately reflect the transient channel structure of the reservoir, and provide a reliable basis for water shut-off management and injection-production optimization.

[0131] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0132] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method of monitoring the production of an oil well based on the concentration of a tracer injected into the well, characterized in that, The method comprises: continuously injecting a background tracer and injecting a pulse tracer according to injection pipeline pressure fluctuations; monitoring the output concentration of the background tracer and the pulse tracer; comparing the output concentration of the pulse tracer with a preset noise threshold, recording a pulse concentration total response sequence; comparing the injection history of the background tracer with the output concentration within a current preset historical neighborhood, constructing a dynamic hydrodynamic baseline parameter set; singular value decomposition of the current pulse concentration total response sequence to reconstruct a single-path concentration response sequence; constructing a hydrodynamic feature set according to each single-path concentration response sequence combined with the dynamic hydrodynamic baseline parameter set; performing pairwise Granger causality test on all single-path concentration response sequences, and constructing a concurrent channeling path system feature set combined with the hydrodynamic feature set; monitoring the total liquid production and the total water production within a current preset monitoring period, and analyzing the invalid circulation amount of the pore type and the fracture type in the total invalid circulation water amount combined with all pulse concentration total response sequences and the corresponding concurrent channeling path system feature set, and generating an output water source composition report.

2. A method of monitoring the production of an oil well based on the concentration of a tracer injected into the well as claimed in claim 1, wherein, The method for obtaining the dynamic hydrodynamic baseline parameter set comprises: deconvolution calculation on the two sequences of the injection history of the background tracer and the output concentration to obtain the average transit time of the background tracer centroid and the time variance of the response curve within a preset historical neighborhood, and input into an exponential moving average filter to output a dynamic baseline transit time and a dynamic baseline dispersion, which constitute the dynamic hydrodynamic baseline parameter set.

3. The method of claim 1, wherein, The method for obtaining the single-path concentration response sequence comprises: converting the pulse concentration total response sequence into a hydrodynamic trajectory matrix through time delay embedding technology; singular value decomposition is performed on the hydrodynamic trajectory matrix, the singular values in the diagonal matrix obtained by decomposition are arranged from large to small, and the singular values are accumulated from large to small until the first preset energy ratio is reached, the singular values participating in accumulation are reserved and the number of concurrent paths is determined; each reserved singular value, the corresponding left singular vector and the right singular vector are combined and reconstructed to obtain the single-path concentration response sequence corresponding to each reserved singular value.

4. The method of claim 2, wherein, The method for obtaining the hydrodynamic feature set comprises: for each single-path concentration response sequence: obtaining a relative velocity index according to the difference between the time from injection to the appearance of the concentration peak and the average transit time; obtaining a relative dispersion index according to the difference between the time variance of the single-path concentration response sequence and the dynamic baseline dispersion; taking the proportion of the square of the singular value in the total sum of the squares of all singular values of the single-path concentration response sequence as the path dominance rate; the relative velocity index, the relative dispersion index and the path dominance rate constitute the hydrodynamic feature set of the single-path concentration response sequence.

5. A method of monitoring the production of an oil well based on the concentration of tracer injected into the well as claimed in claim 4, wherein, The method for obtaining the concurrent channeling path system feature set comprises: performing Granger causality test on each pair of the single-path concentration response sequences to obtain a causality matrix; summing all element values corresponding to a prediction relationship in the causality matrix for each single-path concentration response sequence as a convergence center score; normalizing a sum of all non-diagonal elements in the causality matrix, and taking the normalization result as a system convergence index; constructing the convergence center score, the system convergence index, the total number of the single-path concentration response sequences, and the set of hydrodynamic characteristics of each single-path concentration response sequence into a concurrent channeling path system characteristic set.

6. A method of monitoring the production of an oil well based on the concentration of tracer injected into the well as claimed in claim 5, wherein, The method for generating the produced water source composition report comprises: taking each pulse injection event as a target event in a preset monitoring period; for the target event, fusing the total pulse concentration response sequence and the total produced fluid volume synchronously collected to obtain a total mass of pulse tracer of the target event; multiplying a proportion value of a total mass of the pulse tracer recovered to a total mass of the pulse tracer injected at the time of injection by a total mass of all tracers injected in the process of injecting the pulse tracer to obtain a total ineffective circulation volume; and dividing the total ineffective circulation volume to each single-path concentration response sequence according to the path dominance rate to obtain a single-path leakage volume; when the system convergence index is less than or equal to a preset convergence index threshold value, adding the single-path leakage volume to a pore-type ineffective circulation amount and a fracture-type ineffective circulation amount respectively based on the relative dispersivity index; when the system convergence index is greater than the preset convergence index threshold value, taking the single-path leakage volume corresponding to the maximum relative dispersivity index as a main channel ineffective circulation amount, and adding the single-path leakage volumes corresponding to the remaining relative dispersivity indexes to an upstream tributary ineffective circulation amount; generating a produced water source composition report based on a total produced water volume in a preset monitoring period and various ineffective circulation amounts of all target events.

7. A method of monitoring the production of an oil well based on the concentration of tracer injected into the well as claimed in claim 6, characterised in that, The method for adding the single-path leakage volume to the pore-type ineffective circulation amount and the fracture-type ineffective circulation amount respectively based on the relative dispersivity index comprises: classifying the relative dispersivity index based on a preset dispersivity threshold value, adding the single-path leakage volumes corresponding to the relative dispersivity indexes greater than or equal to the preset dispersivity threshold value to obtain a pore-type ineffective circulation amount, and adding the single-path leakage volumes corresponding to the relative dispersivity indexes less than the preset dispersivity threshold value to obtain a fracture-type ineffective circulation amount.

8. The method of claim 1, wherein, The method for recording the total pulse concentration response sequence comprises: when the produced concentration of the pulse tracer exceeds a preset noise threshold value for the first time, starting to record the produced concentration sequence of the pulse tracer until the produced concentration is lower than the preset noise threshold value and lasts for a preset stable duration, stopping recording, and obtaining a total pulse concentration response sequence.

9. The method of claim 1, wherein, The method for injecting the pulse tracer based on injection pipeline pressure fluctuation comprises: when the injection pipeline pressure reaches a preset pressure proportion of an average pressure of a preset first neighborhood, starting to inject the pulse tracer, and stopping injection after a preset single injection duration.

10. A tracer injection concentration based oil well production monitoring system, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor implements the steps of the oil well production monitoring method based on tracer injection concentration according to any one of claims 1-9 when executing the computer program.

Images