A clastic rock reservoir particle size calculation method and system
By decomposing the T2 spectrum of nuclear magnetic resonance logging using core experiments and blind source separation method, a multi-grain size evaluation model was established, which solved the problem of continuity and accuracy in grain size calculation in complex lithological reservoirs, and realized high-resolution grain size evaluation of the entire well section, supporting fine reservoir evaluation and permeability analysis.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- XINJIANG PETROLEUM ADMINISTRATION BUREAU
- Filing Date
- 2026-07-08
- Publication Date
- 2026-08-04
AI Technical Summary
In existing technologies, the grain size calculation method for complex lithological reservoirs relies on discrete core experiments. Conventional logging curves have poor applicability and cannot achieve continuous and reliable grain size median and grain size percentage parameters throughout the well section. The nuclear magnetic resonance single-spectrum fitting inversion method cannot effectively separate the characteristic signals of different grain sizes.
The proportion of each grain size was obtained through core experiments, the median grain size was determined by laboratory grain size analysis, and the T2 spectrum of nuclear magnetic resonance logging was decomposed by blind source separation method to establish a multi-grain size evaluation model. Correlation analysis was performed between the median grain size and the cumulative porosity of the characteristic relaxation interval of the T2 spectrum of nuclear magnetic resonance logging to achieve continuous calculation of the entire well section.
It enables high-resolution quantitative calculation of the continuous median grain size and the proportion of each grain size in the entire well section of complex lithological reservoirs, improves the continuity and accuracy of grain size parameters, provides key rock structure parameters, and provides a continuous basis for reservoir sedimentary facies analysis, pore structure evaluation and permeability assessment.
Smart Images
Figure CN122507969A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of logging evaluation technology for complex lithological oil and gas reservoirs, and specifically relates to a method and system for calculating the grain size of clastic reservoirs. Background Technology
[0002] In the evaluation and development of complex lithological reservoirs, it is crucial to accurately obtain formation grain size information (such as median grain size and the proportion of each grain size), as this directly controls the reservoir's pore structure, permeability, and fracturing capability.
[0003] The existing technologies for obtaining reservoir grain size mainly include three mainstream technical solutions: core measurement method, conventional logging modeling method, and nuclear magnetic resonance single-spectrum fitting inversion method.
[0004] Among them, core measurement methods include point counting and laser methods. These methods are costly, time-consuming, and only provide discrete point data, making it impossible to achieve continuous evaluation of the entire well section.
[0005] In conventional well logging modeling, although conventional well logging curves (such as natural gamma and resistivity) are sometimes used to estimate grain size parameters, for example, the patent with publication number CN114075965A, in complex reservoirs with complex lithology and strong heterogeneity, the response is affected by multiple factors such as clay minerals and organic matter content, resulting in poor universality and limited accuracy of the evaluation model.
[0006] Nuclear magnetic resonance logging can directly reflect the occurrence state of fluids in pores and pore structure information. Its transverse relaxation time spectrum has an inherent physical correlation with the specific surface area and particle size distribution of rock particles, providing a potential way for continuous particle size evaluation.
[0007] However, existing NMR single-spectrum fitting inversion methods directly use full-spectrum data for modeling, and the characteristic relaxation signals corresponding to different particle sizes are mixed together. They cannot effectively separate and identify the characteristic signals corresponding to different particle size components from the complex, mixed-response NMR T2 spectrum.
[0008] In summary, existing reservoir grain size calculation methods rely on discrete core experiments for grain size evaluation. Conventional logging curves have poor applicability in complex lithology and cannot obtain continuous and reliable grain size median and grain size percentage parameters throughout the well. The nuclear magnetic resonance single-spectrum fitting inversion method cannot achieve high-resolution continuous grain size evaluation throughout the well. Summary of the Invention
[0009] To address the above problems, this invention provides a method for calculating the grain size of clastic rock reservoirs, comprising the following steps: Core experiments were conducted to obtain the proportion of each grain size in the reservoir samples. The median particle size of the reservoir samples was determined using laboratory particle size analysis. The nuclear magnetic resonance logging T2 spectrum of the reservoir was decomposed to obtain the T2 component spectra of multiple fluid components; A multi-particle size evaluation model is established based on the proportion data of each particle size and the T2 component spectra of multiple fluid components. A correlation analysis was conducted between the median particle size and the cumulative porosity in the characteristic relaxation interval of the T2 spectrum from nuclear magnetic resonance logging to establish a median particle size evaluation model. The multi-grain size evaluation model and the grain size median evaluation model were applied to the nuclear magnetic resonance logging data of the entire well section, and the proportion of each grain size and the grain size median were continuously calculated.
[0010] Furthermore, through core experiments and analysis, the proportion of each grain size in the reservoir sample was obtained, including the following steps: Core experiments were conducted to count particles of different sizes within the reservoir sample. Based on the particle size classification criteria, the percentage of particles in each size class was obtained as the proportion data for each size class.
[0011] Furthermore, the median particle size is the particle size value corresponding to the cumulative mass percentage reaching a set value in the cumulative particle size distribution curve.
[0012] Furthermore, the nuclear magnetic resonance logging T2 spectrum of the reservoir is decomposed to obtain T2 component spectra of multiple fluid components, including the following steps: Based on the distribution characteristics of the T2 spectrum of nuclear magnetic resonance logging in the reservoir, the fluid component types are determined; based on the number of fluid component types in the reservoir, the number of sub-spectrums is determined; the nuclear magnetic resonance logging T2 spectrum is decomposed into T2 relaxation feature matrices containing different fluid components and volume matrices containing different fluid components using the blind source separation method; based on the T2 relaxation features of different fluid components and the volume matrices containing different fluid components, multiple fluid component T2 component spectra are obtained.
[0013] Furthermore, the nuclear magnetic resonance logging T2 spectrum is decomposed into T2 relaxation characteristic matrices containing different fluid components and volume matrices containing different fluid components using the blind source separation method, including the following steps: The T2 spectra from nuclear magnetic resonance logging at m depths in the reservoir are arranged and laid out to construct the original nuclear magnetic resonance data matrix V with dimensions of n rows and m columns. n×m Where n represents the number of relaxation time points in a single nuclear magnetic resonance logging T2 spectrum; Based on the number of fluid component types 'a', the original NMR data matrix V n×m Decomposed into the initial T2 relaxor characteristic matrix W n×a and volume matrix H a×mSet an upper limit for the number of iterations and an error threshold; iterate the T2 relaxation feature matrix and the volume matrix H sequentially according to the iterative update formula of the T2 relaxation feature matrix and the volume matrix H. The iteration ends when the number of iterations reaches the upper limit or when the deviation between the original NMR data matrix and the product of the iterated T2 relaxation feature matrix and the volume matrix is less than the error threshold. Based on the number of fluid components obtained in the previous step, use non-negative matrix decomposition to obtain the T2 relaxation feature matrix and the volume matrix containing a kinds of fluid components.
[0014] Furthermore, based on the proportion data of each particle size and the T2 component spectra of multiple fluid components, a multi-particle size evaluation model is established, including the following steps: The proportion data of each particle size is matched with the T2 spectrum of nuclear magnetic resonance logging at the corresponding depth point. The proportion data of each particle size is matched with the T2 component spectrum of multiple fluid components. The porosity of the T2 component spectrum of each fluid component is calculated. A multi-particle size evaluation model is established based on the relationship between the porosity of the T2 component spectrum of each fluid component and the proportion of each particle size.
[0015] Furthermore, the multi-grain size evaluation model includes a mud-grade evaluation model, a silt-grade evaluation model, a fine-sand-grade evaluation model, a medium-sand-grade evaluation model, and a coarse-sand and gravel evaluation model.
[0016] Furthermore, a correlation analysis was conducted between the median grain size and the cumulative porosity within the characteristic relaxation interval of the T2 spectrum from nuclear magnetic resonance logging to establish a median grain size evaluation model, including the following steps: By comparing and performing correlation analysis between the median grain size and the T2 spectrum of nuclear magnetic resonance logging at the corresponding depth, the characteristic relaxation intervals in the T2 spectrum that show a linear relationship with the median grain size are determined, and the cumulative porosity of the characteristic relaxation intervals is calculated. Based on the cumulative porosity of the characteristic relaxation intervals and the median grain size, a grain size evaluation model is established.
[0017] Furthermore, the granularity classification criteria include: Particles smaller than 15.6 μm are classified as mud-grade particles; particles between 15.6 and 62.5 μm are classified as silt-grade particles; particles between 62.5 and 250 μm are classified as fine sand-grade particles; particles between 250 and 500 μm are classified as medium sand-grade particles; and particles larger than 500 μm are classified as coarse sand and gravel-grade particles.
[0018] This invention also provides a grain size calculation system for clastic rock reservoirs, comprising: The core experimental analysis unit is used to obtain the proportion data of each grain size in the reservoir sample through core experimental analysis; and to determine the median grain size of the reservoir sample using laboratory grain size analysis. The nuclear magnetic resonance spectroscopy decomposition unit is used to decompose the nuclear magnetic resonance logging T2 spectrum of the reservoir to obtain the T2 component spectra of multiple fluid components; The first model building unit is used to establish a multi-level evaluation model based on the proportion data of each particle size and the T2 component spectrum of multiple fluid components. The second model construction unit is used to perform correlation analysis between the median particle size and the cumulative porosity of the characteristic relaxation interval of the T2 spectrum of nuclear magnetic resonance logging, and to establish a median particle size evaluation model. The particle size calculation unit is used to apply the multi-particle size evaluation model and the particle size median evaluation model to the nuclear magnetic resonance logging data of the entire well section, and continuously calculate the proportion of each particle size and the particle size median.
[0019] Furthermore, the first model building unit is specifically used for: The proportion data of each particle size is matched with the T2 spectrum of nuclear magnetic resonance logging at the corresponding depth point. The proportion data of each particle size is matched with the T2 component spectrum of multiple fluid components. The porosity of the T2 component spectrum of each fluid component is calculated. A multi-particle size evaluation model is established based on the relationship between the porosity of the T2 component spectrum of each fluid component and the proportion of each particle size.
[0020] Furthermore, the second model building unit is specifically used for: By comparing and performing correlation analysis between the median grain size and the T2 spectrum of nuclear magnetic resonance logging at the corresponding depth, the characteristic relaxation intervals in the T2 spectrum that show a linear relationship with the median grain size are determined, and the cumulative porosity of the characteristic relaxation intervals is calculated. Based on the cumulative porosity of the characteristic relaxation intervals and the median grain size, a grain size evaluation model is established.
[0021] The beneficial effects of this invention are: This invention utilizes the T2 spectrum of nuclear magnetic resonance (NMR) logging in reservoirs, decomposing the NMR T2 spectrum into different fluid components. Based on the proportion of each particle size and the T2 component spectrum of the fluid components, a multi-particle size evaluation model is established. Furthermore, based on the median particle size and the cumulative porosity within the characteristic relaxation interval of the NMR T2 spectrum, a median particle size evaluation model is established. Through these two models, the continuous quantitative calculation of the median particle size and the proportion of each particle size (mud, silt, sand, etc.) throughout the entire well section is successfully achieved, clearly revealing the fine distribution of the reservoir's particle size structure vertically. Through core calibration verification, this invention significantly improves the continuity and accuracy of particle size parameter evaluation, effectively solving the problems of traditional methods relying on discrete cores and high ambiguity in conventional logging. It can provide key and continuous rock structure parameters for reservoir sedimentary facies analysis, pore structure evaluation, permeability assessment, and engineering "sweet spot" selection. The entire process is efficient and convenient, facilitating rapid decision-making and large-scale application in oilfield exploration and development. It has outstanding universality and promotional value for evaluating complex lithological reservoirs with strong heterogeneity.
[0022] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description and the drawings. Attached Figure Description
[0023] To more clearly illustrate the technical solutions 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 some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0024] Figure 1 A flowchart illustrating a method for calculating the grain size of clastic rock reservoirs according to an embodiment of the present invention is shown. Figure 2 A schematic diagram of rock particle size statistics according to an embodiment of the present invention is shown; Figure 3 A schematic diagram of the T2 spectrum results of blind source separation decomposition nuclear magnetic resonance logging according to an embodiment of the present invention is shown; Figure 4 A schematic diagram of a mud grade evaluation model according to an embodiment of the present invention is shown; Figure 5 A schematic diagram of a silt grade evaluation model according to an embodiment of the present invention is shown; Figure 6 A schematic diagram of a fine sand grade evaluation model according to an embodiment of the present invention is shown; Figure 7 A schematic diagram of a medium sand grade evaluation model according to an embodiment of the present invention is shown; Figure 8 A schematic diagram of a coarse sand and gravel grade evaluation model according to an embodiment of the present invention is shown; Figure 9 A schematic diagram of a granularity median evaluation model according to an embodiment of the present invention is shown; Figure 10 A schematic diagram of a grain size calculation system for clastic rock reservoirs according to an embodiment of the present invention is shown. Figure 11 A graph showing the particle size evaluation results of well 207 according to an embodiment of the present invention is displayed. Detailed Implementation
[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.
[0026] It should be noted that the terms "first," "second," etc., used in this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate for the embodiments of this application described herein.
[0027] This invention provides a method and system for calculating the grain size of clastic reservoirs, addressing the problems in existing technologies where grain size evaluation relies on discrete core experiments, conventional logging curves have poor applicability in complex lithologies, and continuous and reliable grain size median and grain size percentage parameters cannot be obtained throughout the well. By combining nuclear magnetic resonance logging response mechanism with blind source separation algorithm, this invention achieves high-resolution, continuous, and quantitative calculation of reservoir grain size median and the percentage of each grain size (mud, silt, sand, etc.), eliminating reliance on conventional logging curves and providing key and continuous geological parameter basis for fine reservoir evaluation, permeability analysis, and engineering "sweet spot" selection.
[0028] like Figure 1 As shown, a method for calculating the grain size of clastic rock reservoirs includes the following steps: S1. Obtain the proportion data of each particle size in the reservoir sample through core experiments.
[0029] For example, by analyzing core samples, the number of particles of different sizes in the reservoir sample is counted, and the percentage of particles in each size class is obtained according to the size classification standard, which is used as the proportion data of each size class.
[0030] The core experimental analysis determines the particle size distribution, including but not limited to: thin section image analysis (using point counting or digital image processing techniques to count the proportion of particles in each size), laser diffraction particle size analysis, and sedimentation particle size analysis. The accurate particle size distribution data obtained in this step will serve as the key experimental calibration basis for subsequently establishing a quantitative relationship model between nuclear magnetic resonance response and particle size composition.
[0031] For example, the particle size classification criteria are shown in Table 1. By measuring reservoir samples, the volume or number percentages of mud-grade particles with a particle size less than 15.6 μm, silt-grade particles with a particle size between 15.6 and 62.5 μm, fine sand-grade particles with a particle size between 62.5 and 250 μm, medium sand-grade particles with a particle size between 250 and 500 μm, and coarse sand and gravel-grade particles with a particle size greater than 500 μm are statistically calculated.
[0032] Table 1
[0033] Statistical results are as follows Figure 2 As shown, the horizontal axis represents the particle size range in μm, corresponding to the five particle size classification standards mentioned above, and the vertical axis represents the volume percentage (mass percentage, %) of each particle size class.
[0034] Figure 2 In this sample, mud-grade particles accounted for 3.69%, and silt-grade particles accounted for 81.49%, representing the dominant particle size distribution. The main particle composition was silt. Fine sand accounted for 11.17%, medium sand for 2.70%, coarse sand for 0.69%, and gravel for 0.27%. The median particle size distribution was D. 50 =29.10 falls within the 15.6~62.5μm silt range, which matches the characteristics of the main component of this sample, where silt particles account for more than 80%.
[0035] S2. Using laboratory particle size analysis, determine the median particle size (D) of the reservoir samples. 50 ).
[0036] Median particle size (D) 50 The particle size distribution (PMD) is a key parameter characterizing the coarseness of rock particles. It is defined as the particle size value corresponding to the cumulative mass percentage in the cumulative particle size distribution curve reaching a set value (e.g., 50%).
[0037] For example, the median particle size (D) of reservoir samples can be obtained through laboratory particle size analysis. 50 Specific methods include, but are not limited to: obtaining the median grain size through image statistical analysis using the MapScan mineral scanning experimental system; obtaining the median grain size D from core samples of the same depth or layer through image statistical analysis using mineral scanning systems such as QEMSCAN, or direct measurement using equipment such as laser grain size analyzers. 50 .
[0038] Median particle size (D) 50 This is used for subsequent correlation analysis with nuclear magnetic resonance logging information to establish and calibrate a continuous calculation model for grain size median using nuclear magnetic resonance logging T2 spectrum, as in step S5.
[0039] S3. Decompose the T2 spectrum of the reservoir using nuclear magnetic resonance logging to obtain T2 component spectra of multiple fluid components.
[0040] like Figure 3 As shown, this is a schematic diagram of the blind source separation and decomposition of the nuclear magnetic resonance logging T2 spectrum results in an embodiment of the present invention. The nuclear magnetic resonance logging T2 spectrum is the sum of all pore fluid signals in the reservoir. It is a mixture of relaxation signals of various fluids (such as bound water, capillary bound water, and free fluid) in pores of different sizes. The mixed nuclear magnetic T2 spectrum can be decomposed into multiple independent fluid component spectra. Each component corresponds to a pore fluid or pore structure with specific relaxation characteristics.
[0041] For example, step S3 specifically includes the following steps: determining the fluid component type based on the distribution characteristics of the reservoir's nuclear magnetic resonance logging T2 spectrum; determining the number of sub-spectrums (usually 6 to 7) based on the number of fluid component types in the reservoir; decomposing the nuclear magnetic resonance logging T2 spectrum into T2 relaxation feature matrices containing different fluid components and volume matrices containing different fluid components using the blind source separation method; and obtaining multiple fluid component T2 component spectra based on the T2 relaxation features of different fluid components and the volume matrices containing different fluid components.
[0042] For example, based on the distribution characteristics of the T2 spectrum of nuclear magnetic resonance logging in the reservoir, seven types of fluid components were identified. Based on the number of fluid component types in the reservoir, the T2 spectrum of nuclear magnetic resonance logging was decomposed using the blind source separation method. The T2 spectrum of nuclear magnetic resonance logging was decomposed into two matrices. The first matrix contains the T2 relaxation characteristics of different fluid components, and the second matrix contains the volume of each fluid component.
[0043] For example, using the blind source separation method to decompose the T2 spectrum of nuclear magnetic resonance logging into T2 relaxation features containing different fluid components, and a volume matrix containing different fluid components, includes the following steps: The T2 spectra from nuclear magnetic resonance logging at m depths in the reservoir are arranged and laid out to construct the original nuclear magnetic resonance data matrix V with dimensions of n rows and m columns. n×m In this matrix, the number of rows n represents the number of relaxation time points for a single nuclear magnetic resonance logging T2 spectrum, and the number of columns m represents the number of logging points at the corresponding logging depth.
[0044] Based on the number of fluid component types 'a', the original NMR data matrix V n×m Decomposed into the initial T2 relaxor characteristic matrix W n×a and volume matrix H a×mSecondly, set an upper limit for the number of iterations and an error threshold. Finally, according to the iterative update formula (1) for the T2 relaxation feature matrix W and the iterative update formula (2) for the volume matrix H, iterate the T2 relaxation feature matrix and the volume matrix sequentially. The iteration ends when the number of iterations reaches the upper limit, or when the deviation between the original NMR data matrix and the product of the iterated T2 relaxation feature matrix and the volume matrix is less than the error threshold. Based on the number of fluid components obtained in the previous step, the T2 relaxation feature matrix W and the volume matrix H containing a kinds of fluid components are obtained by non-negative matrix decomposition.
[0045] (1) (2) In the formula, The original NMR T2 spectrum data matrix ( ); For the coefficient matrix ( ), including the proportion of each fluid component at each depth point; Basis matrix ( ),Include T2 relaxation characteristic spectra of various fluid components; The number of fluid components; This represents the number of iterations. The row number of the matrix; The column index of the matrix. This is the matrix transpose operator. For the first The volume matrix H of the next iteration Line number The old elements of the column, For the first +1 iteration update of the volume matrix H Line number New elements in the column, Indicates the first The volume matrix H of the next iteration, Indicates the first The T2 relaxation characteristic matrix W of the next iteration; For the first The T2 relaxation feature matrix W of the next iteration Line number The old elements of the column, For the first +1 iteration update of the T2 relaxation feature matrix W Line number New elements in the column.
[0046] Based on the T2 relaxation characteristics of different fluid components and the volume matrix containing different fluid components, the T2 component spectra of multiple fluid components are obtained, including the following steps: After obtaining the T2 relaxation characteristic matrix W and the volume matrix H, each column of the W matrix is used as a standard T2 relaxation characteristic spectrum of a fluid component, and the corresponding row in the H matrix is used as the volume contribution coefficient of that fluid component along the depth direction. For the , r For fluid-like components, the T2 component spectral matrix at all depth points can be derived from the W matrix. j Column and H matrix i The T2 spectrum is obtained by multiplying the rows. For any depth point, after calculating the T2 component spectrum of all fluid components in sequence, the nuclear magnetic resonance logging T2 spectrum can be decomposed into several T2 component spectra with different fluid properties, and the sum of each component spectrum is approximately equal to the original T2 spectrum.
[0047] S4. Based on the proportion data of each particle size and the T2 component spectra of multiple fluid components, establish a multi-particle size evaluation model.
[0048] Based on the surface relaxation mechanism of nuclear magnetic resonance, the proportion of each grain size in the core should correspond to the T2 spectrum of each fluid component.
[0049] For example, step S4 specifically includes the following steps: accurately matching the proportion data of each grain size with the T2 spectrum of the nuclear magnetic resonance logging at the corresponding depth point. Considering the correlation between the nuclear magnetic resonance surface relaxation mechanism and grain size, i.e., the complex pore size distribution in complex lithological reservoirs is mainly controlled by the grain size composition of the rock, the proportion data of each grain size is matched with the T2 component spectra of multiple fluid components, and the porosity of each fluid component's T2 component spectrum is calculated. Based on the relationship between the porosity of each fluid component's T2 component spectrum and the proportion of each grain size, a multi-grain size evaluation model is established.
[0050] Among them, the multi-grain size evaluation model includes mud-grade evaluation model, silt-grade evaluation model, fine sand-grade evaluation model, medium sand-grade evaluation model, and coarse sand and gravel evaluation model.
[0051] For example, the mud-grade (particle size <15.6 μm) evaluation model: extremely fine particles, large specific surface area, poor pore throat connectivity, and high bound water content. Due to surface relaxation, its corresponding transverse NMR relaxation time (T2) is extremely short. Therefore, the porosity Por of the T2 component spectrum of fluid components with the main peak located in the 0.3–3 ms range is used. a S with mud grade a The evaluation model is established, and the formula is as follows: S a =k a Por a +b a (3) In the formula, k is the linear regression coefficient of the mud grade evaluation model. a b a Calibration needs to be performed through regional experiments, such as Figure 4As shown, k a =8.4363, b a =0.109, coefficient of determination (goodness of fit) R 2 =0.6858, Figure 4 The horizontal axis represents porosity (0.3~3ms), the vertical axis represents the clay content, the orange dots represent measured data points, each point represents the measured value of porosity and clay content of a reservoir sample, and the red dashed line represents the linear regression fitting line, with the corresponding formula being y=8.4363x+0.109.
[0052] For example, the evaluation model for silty sand (particle size 15.6–62.5 μm) shows relatively coarse particles, improved pore connectivity, weakened surface relaxation intensity, and a significantly longer transverse nuclear magnetic resonance relaxation time (T2) response compared to mud-grade sand. The porosity Por of the T2 component spectrum of the fluid component with the main peak located in the 3–50 ms range is used. b With the proportion of silt S b The evaluation model is established, and the formula is as follows: S b =k b Por b +b b (4) In the formula, k is the linear regression coefficient of the silt grade evaluation model. b b b Calibration needs to be performed through regional experiments, such as Figure 5 As shown, k b =11.122, b b =0.162, coefficient of determination (goodness of fit) R 2 =0.5814, Figure 5 The horizontal axis represents porosity (3~50ms), the vertical axis represents the silt content, the purple dots represent measured data points, each point represents the measured value of porosity and silt content of a reservoir sample, and the red dashed line represents the linear regression fitting line, with the corresponding formula being y=11.122x+0.162.
[0053] Evaluation model for fine sand (particle size 62.5–250 μm): Further increasing the particle size leads to a more open pore structure, an increased proportion of free fluid, a further reduction in the influence of surface relaxation, and a significant increase in the transverse relaxation time (T2) of nuclear magnetic resonance (NMR). Its characteristic response is concentrated in the 50–120 ms range, within which the porosity Por of the fluid component T2 component spectrum is [value missing]. c With the proportion of fine sand S c The evaluation model is established, and the formula is as follows: S c =k c Por c +b c(5) In the formula, k is the linear regression coefficient of the fine sand grade evaluation model. c b c Calibration needs to be performed through regional experiments, such as Figure 6 As shown, k c =18.123, b c =0.0681, coefficient of determination (goodness of fit) R 2 =0.6685, Figure 6 In the diagram, the horizontal axis represents porosity (50~120ms), the vertical axis represents the proportion of fine sand, the blue dots represent measured data points, each point represents the measured value of porosity and fine sand proportion of a reservoir sample, and the red dashed line represents the linear regression fitting line, with the corresponding formula being y=18.123x+0.0681.
[0054] Evaluation model for medium-sized sand (particle size 250–500 μm): Enhanced fluid mobility and further prolonged transverse relaxation time (T2) on nuclear magnetic resonance (NMR). The porosity P of the T2 component spectrum of the fluid component with the main peak located in the 120–300 ms range was used. d With the proportion of medium sand S d The evaluation model is established, and the formula is as follows: S d =k d Por d +b d (6) In the formula, k is the linear regression coefficient of the medium sand grade evaluation model. d b d Calibration needs to be performed through regional experiments, such as Figure 7 As shown, k d =14.9, b d =0.0358, coefficient of determination (goodness of fit) R 2 =0.7306, Figure 7 In the diagram, the horizontal axis represents porosity (120~300ms), the vertical axis represents the proportion of medium sand, the green dots represent measured data points, each point corresponds to the measured values of porosity and medium sand proportion of a reservoir sample, and the red dashed line represents the linear regression fitting line, corresponding to the formula y=14.9x+0.0358.
[0055] Evaluation model for coarse sand and gravel (particle size > 500 μm): Typically very low in mixed-phase shale reservoirs, usually occurring as isolated interlayers or laminae. The transverse relaxation time (T2) response of nuclear magnetic resonance (NMR) is located in an ultra-long relaxation range > 300 ms, with a weak signal easily affected by wellbore fluids. The porosity Por of the fluid component T2 component spectrum with the main peak located in the 300–800 ms range is used. e The ratio of coarse sand to gravel (S) e The evaluation model is established, and the formula is as follows: S e =k e Por e +b e (7) In the formula, k is the linear regression coefficient of the coarse sand and gravel evaluation model. e b e Calibration needs to be performed through regional experiments, such as Figure 8 As shown, k e =19.444, b e =0.0427, coefficient of determination (goodness of fit) R 2 =0.6635, Figure 8 In the diagram, the horizontal axis represents porosity (300~800ms), the vertical axis represents the proportion of coarse sand and gravel, the blue dots represent measured data points, each point corresponds to the measured value of porosity and the proportion of coarse sand and gravel of a reservoir sample, and the blue dashed line is the linear regression fitting line, corresponding to the formula y=19.444x+0.0427.
[0056] S5. Correlation analysis was performed between the median particle size and the cumulative porosity of the characteristic relaxation interval of the T2 spectrum of nuclear magnetic resonance logging to establish a median particle size evaluation model.
[0057] like Figure 9 This is a diagram illustrating the median grain size evaluation model according to an embodiment of the present invention. In reservoirs dominated by surface relaxation, the grain size distribution of the rock directly controls its pore radius distribution. An increase in the median grain size generally indicates an increase in the coarse-grained component of the rock, leading to an overall shift in pore radius towards larger sizes.
[0058] For example, step S5 specifically includes the following steps: performing a detailed comparison and correlation analysis between the median grain size and the corresponding depth point's nuclear magnetic resonance logging T2 spectrum, to determine the grain size D in the nuclear magnetic resonance logging T2 spectrum. 50 The characteristic relaxation intervals exhibiting linear relationships are identified, and the cumulative porosity of these intervals is calculated. Based on the cumulative porosity of the characteristic relaxation intervals and the median particle size, a median particle size evaluation model is established.
[0059] For example, a detailed comparison and correlation analysis of the median grain size with the T2 spectra of nuclear magnetic resonance logging at the corresponding depth points revealed that the cumulative porosity within a specific T2 interval of 2–60 ms is related to the median grain size D. 50 It exhibits the best linear correlation. Based on core experimental data and calibration data, a grain size median evaluation model is established, as shown in the following formula: D 50 =kPor 2~60 +c(8) In the formula, the coefficients k and c need to be calibrated through regional experiments, such as... Figure 9As shown, k=17.854, c=1.0168, R 2 =0.8516, Por 2~60 This represents the cumulative porosity within the specific T2 interval of 2 to 60 ms. Figure 9 In the figure, the horizontal axis represents porosity (%) from 2 to 60 ms, the vertical axis represents the median particle size (μm), the red dots represent the measured data points, and each point corresponds to the measured values of porosity and median particle size of a reservoir sample. The black dashed line is the linear regression fitting line, corresponding to the formula y=17.854x+1.0168.
[0060] S6. Apply the multi-grain size evaluation model and the grain size median evaluation model to the nuclear magnetic resonance logging data of the entire well section, and continuously calculate the proportion of each grain size and the grain size median to complete the reservoir grain size evaluation.
[0061] After completing the multi-granularity evaluation model and granularity median (D) 50 After the evaluation model is constructed and calibrated, the above model system is integrated and applied to the nuclear magnetic resonance logging data of the entire well section of the target well. High-resolution grain size median curves and the proportion curves of each grain size are continuously calculated to achieve continuous, high-resolution, and multi-parameter comprehensive evaluation of the grain size characteristics of complex lithological reservoirs.
[0062] Based on the above-mentioned methods for calculating the grain size of clastic reservoirs, such as Figure 10 As shown in the figure, this embodiment of the invention also provides a grain size calculation system for clastic rock reservoirs, including a core experimental analysis unit, a nuclear magnetic resonance spectroscopy decomposition unit, a first model construction unit, a second model construction unit, and a grain size calculation unit.
[0063] The core experimental analysis unit is used to obtain the proportion data of each grain size in the reservoir sample through core experimental analysis; and to determine the median grain size of the reservoir sample using laboratory grain size analysis.
[0064] The nuclear magnetic resonance (NMR) spectroscopy decomposition unit is used to decompose the NMR logging T2 spectrum of the reservoir to obtain multiple fluid component T2 component spectra.
[0065] The first model building unit is used to establish a multi-size evaluation model based on the proportion data of each particle size and the T2 component spectra of multiple fluid components. The second model building unit is used to perform correlation analysis between the median particle size and the cumulative porosity of the characteristic relaxation interval of the nuclear magnetic resonance logging T2 spectrum to establish a median particle size evaluation model.
[0066] The particle size calculation unit is used to apply the multi-particle size evaluation model and the particle size median evaluation model to the nuclear magnetic resonance logging data of the entire well section, and continuously calculate the proportion of each particle size and the particle size median.
[0067] The grain size calculation method and system for clastic rock reservoirs according to embodiments of the present invention were applied to Well 207, and the application effect of the present invention was verified based on the results of core grain size and particle size experiments. Figure 11 As shown, Figure 11 The image shows the particle size evaluation results for Well 207 (the eighth channel is the result of the mud grade ratio calculation; the curve represents the result of the mud grade ratio calculated using the T2 spectrum of nuclear magnetic resonance logging; and the dots represent the mud grade ratio results determined by core experiments).
[0068] The ninth channel is the result channel for calculating the proportion of silt grade. The curve represents the result of calculating the proportion of silt grade using nuclear magnetic resonance T2 spectrum logging data, and the dots represent the result of determining the proportion of silt grade by core experiments.
[0069] The tenth channel is the result channel for calculating the proportion of fine sand. The curve represents the result of calculating the proportion of fine sand using nuclear magnetic resonance T2 spectrum logging data, and the dots represent the result of determining the proportion of fine sand by core experiments.
[0070] The eleventh channel is the result channel for calculating the proportion of medium sand. The curve represents the result of calculating the proportion of medium sand using nuclear magnetic resonance T2 spectrum logging data, and the dots represent the result of determining the proportion of medium sand by core experiments.
[0071] The twelfth channel is the result channel for calculating the proportion of coarse sand and above. The curve represents the result of calculating the proportion of coarse sand and above using nuclear magnetic resonance T2 spectrum logging data, and the dots represent the results of determining the proportion of coarse sand and gravel grades by core experiments.
[0072] The thirteenth channel shows the results of the median grain size calculation. The curve represents the result of calculating the median grain size using nuclear magnetic resonance T2 spectral logging data, and the dots represent the results of determining the median grain size through core experiments. It can be seen that the grain size and median grain size determined experimentally in the figure are basically consistent with the grain size proportion and median grain size calculated by the method of this invention, indicating that this invention can accurately evaluate the grain size of complex lithological reservoirs.
[0073] In summary, the grain size calculation method and system for clastic rock reservoirs of this invention first obtains the grain size proportion and median grain size (D) at discrete depth points through core experiments. 50 The data was analyzed; then, the blind source separation algorithm was used to decompose the T2 spectrum of the nuclear magnetic resonance logging of the entire well section, separating the mixed relaxation signal into multiple independent fluid component spectra representing different pore size systems; based on this, combined with the surface relaxation physical mechanism, the characteristic relaxation time intervals of each fluid component spectrum were correlated with specific particle sizes (mud, silt, fine sand, medium sand, coarse sand and gravel), and calibrated using core experimental data to establish quantitative linear evaluation models for the proportion of each particle size and the median particle size; finally, these models were integrated and applied to process the nuclear magnetic resonance data of the entire well section, and high-resolution median particle size curves and proportion curves of each particle size were continuously calculated.
[0074] This invention boasts a rigorous theoretical foundation and a clearly defined physical mechanism (surface relaxation-dominated), effectively addressing two major challenges in grain size evaluation of complex lithological reservoirs: discretized core data and the difficulty in interpreting conventional logging data due to its ambiguity. By introducing a blind source separation algorithm, the T2 spectrum from nuclear magnetic resonance (NMR) logging is decomposed into different fluid components, and a quantitative conversion model between component relaxation time and rock grain size (mud, silt, sand, etc.) is established. This enables, for the first time, the calculation of continuous grain size parameters (median grain size and percentage of each grain size) across the entire well section based on single NMR logging data. In practical applications, this method can produce high-resolution continuous grain size profiles, providing key parameters for reservoir sedimentary microfacies characterization, pore structure evaluation, permeability analysis, and the selection of optimal "sweet spots" in engineering, strongly supporting the refined evaluation and efficient development of complex oil and gas reservoirs. With its clear principles and standardized processing procedures, this method combines theoretical rigor with engineering convenience, making it of significant value for widespread application.
[0075] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for calculating the grain size of clastic rock reservoirs, characterized in that, Includes the following steps: Core experiments were conducted to obtain the proportion of each grain size in the reservoir samples. The median particle size of the reservoir samples was determined using laboratory particle size analysis. The nuclear magnetic resonance logging T2 spectrum of the reservoir was decomposed to obtain the T2 component spectra of multiple fluid components; A multi-particle size evaluation model is established based on the proportion data of each particle size and the T2 component spectra of multiple fluid components. A correlation analysis was conducted between the median particle size and the cumulative porosity in the characteristic relaxation interval of the T2 spectrum from nuclear magnetic resonance logging to establish a median particle size evaluation model. The multi-grain size evaluation model and the grain size median evaluation model were applied to the nuclear magnetic resonance logging data of the entire well section, and the proportion of each grain size and the grain size median were continuously calculated.
2. The method for calculating the grain size of clastic rock reservoirs according to claim 1, characterized in that, Core experiments were conducted to obtain the proportion of each grain size in the reservoir sample, including the following steps: Core experiments were conducted to count particles of different sizes within the reservoir sample. Based on the particle size classification criteria, the percentage of particles in each size class was obtained as the proportion data for each size class.
3. The method for calculating the grain size of clastic rock reservoirs according to claim 1, characterized in that, The median particle size is the particle size value corresponding to the cumulative mass percentage reaching a set value in the cumulative particle size distribution curve.
4. The method for calculating the grain size of clastic rock reservoirs according to claim 1, characterized in that, The nuclear magnetic resonance logging T2 spectrum of the reservoir is decomposed to obtain T2 component spectra of multiple fluid components, including the following steps: Based on the distribution characteristics of the T2 spectrum of nuclear magnetic resonance logging in the reservoir, the fluid component types are determined; based on the number of fluid component types in the reservoir, the number of sub-spectrums is determined; the nuclear magnetic resonance logging T2 spectrum is decomposed into T2 relaxation feature matrices containing different fluid components and volume matrices containing different fluid components using the blind source separation method; based on the T2 relaxation features of different fluid components and the volume matrices containing different fluid components, multiple fluid component T2 component spectra are obtained.
5. The method for calculating the grain size of clastic rock reservoirs according to claim 4, characterized in that, The T2 spectrum of nuclear magnetic resonance logging is decomposed into T2 relaxation characteristic matrices containing different fluid components and volume matrices containing different fluid components using the blind source separation method, including the following steps: The T2 spectrum of the nuclear magnetic resonance logging of m depth points of a reservoir is arranged to obtain an original nuclear magnetic data matrix V with n rows and m columns n×m wherein n represents the number of relaxation time points of a single T2 spectrum of the nuclear magnetic resonance logging. Based on the number of fluid component types 'a', the original NMR data matrix V n×m Decomposed into the initial T2 relaxor characteristic matrix W n×a and volume matrix H a×m Set an upper limit for the number of iterations and an error threshold; iterate the T2 relaxation feature matrix and the volume matrix H sequentially according to the iterative update formula of the T2 relaxation feature matrix and the volume matrix H. The iteration ends when the number of iterations reaches the upper limit or when the deviation between the original NMR data matrix and the product of the iterated T2 relaxation feature matrix and the volume matrix is less than the error threshold. Based on the number of fluid components obtained in the previous step, use non-negative matrix decomposition to obtain the T2 relaxation feature matrix and the volume matrix containing a kinds of fluid components.
6. The method for calculating the grain size of clastic rock reservoirs according to any one of claims 1-5, characterized in that, Based on the proportion data of each particle size and the T2 component spectra of multiple fluid components, a multi-particle size evaluation model is established, including the following steps: The proportion data of each particle size is matched with the T2 spectrum of nuclear magnetic resonance logging at the corresponding depth point. The proportion data of each particle size is matched with the T2 component spectrum of multiple fluid components. The porosity of the T2 component spectrum of each fluid component is calculated. A multi-particle size evaluation model is established based on the relationship between the porosity of the T2 component spectrum of each fluid component and the proportion of each particle size.
7. The method for calculating the grain size of clastic rock reservoirs according to claim 6, characterized in that, The multi-grain size evaluation model includes mud-scale evaluation model, silt-scale evaluation model, fine sand-scale evaluation model, medium sand-scale evaluation model, and coarse sand and gravel evaluation model.
8. The method for calculating the grain size of clastic rock reservoirs according to any one of claims 1-5, characterized in that, A correlation analysis was performed between the median grain size and the cumulative porosity within the characteristic relaxation interval of the T2 spectrum from nuclear magnetic resonance logging to establish a median grain size evaluation model, including the following steps: By comparing and performing correlation analysis between the median grain size and the T2 spectrum of nuclear magnetic resonance logging at the corresponding depth, the characteristic relaxation intervals in the T2 spectrum that show a linear relationship with the median grain size are determined, and the cumulative porosity of the characteristic relaxation intervals is calculated. Based on the cumulative porosity of the characteristic relaxation intervals and the median grain size, a grain size evaluation model is established.
9. The method for calculating the grain size of clastic rock reservoirs according to claim 1, characterized in that, Grain size classification criteria include: Particles smaller than 15.6 μm are classified as mud-grade particles; particles between 15.6 and 62.5 μm are classified as silt-grade particles; particles between 62.5 and 250 μm are classified as fine sand-grade particles; particles between 250 and 500 μm are classified as medium sand-grade particles; and particles larger than 500 μm are classified as coarse sand and gravel-grade particles.
10. A grain size calculation system for clastic rock reservoirs, characterized in that, include: The core experiment analysis unit is used to obtain the proportion data of each grain size in the reservoir sample through core experiment analysis; The median particle size of the reservoir samples was determined using laboratory particle size analysis. The nuclear magnetic resonance spectroscopy decomposition unit is used to decompose the nuclear magnetic resonance logging T2 spectrum of the reservoir to obtain the T2 component spectra of multiple fluid components; The first model building unit is used to establish a multi-level evaluation model based on the proportion data of each particle size and the T2 component spectrum of multiple fluid components. The second model construction unit is used to perform correlation analysis between the median particle size and the cumulative porosity of the characteristic relaxation interval of the T2 spectrum of nuclear magnetic resonance logging, and to establish a median particle size evaluation model. The particle size calculation unit is used to apply the multi-particle size evaluation model and the particle size median evaluation model to the nuclear magnetic resonance logging data of the entire well section, and continuously calculate the proportion of each particle size and the particle size median.
11. The grain size calculation system for clastic rock reservoirs according to claim 10, characterized in that, The first model building unit is specifically used for: The proportion data of each particle size is matched with the T2 spectrum of nuclear magnetic resonance logging at the corresponding depth point. The proportion data of each particle size is matched with the T2 component spectrum of multiple fluid components. The porosity of the T2 component spectrum of each fluid component is calculated. A multi-particle size evaluation model is established based on the relationship between the porosity of the T2 component spectrum of each fluid component and the proportion of each particle size.
12. The grain size calculation system for clastic rock reservoirs according to claim 10, characterized in that, The second model building unit is specifically used for: By comparing and performing correlation analysis between the median grain size and the T2 spectrum of nuclear magnetic resonance logging at the corresponding depth, the characteristic relaxation intervals in the T2 spectrum that show a linear relationship with the median grain size are determined, and the cumulative porosity of the characteristic relaxation intervals is calculated. Based on the cumulative porosity of the characteristic relaxation intervals and the median grain size, a grain size evaluation model is established.