CN116840906A - Uncertainty evaluation method for reservoir physical property parameter prediction - Google Patents
Uncertainty evaluation method for reservoir physical property parameter prediction Download PDFInfo
- Publication number
- CN116840906A CN116840906A CN202210288063.6A CN202210288063A CN116840906A CN 116840906 A CN116840906 A CN 116840906A CN 202210288063 A CN202210288063 A CN 202210288063A CN 116840906 A CN116840906 A CN 116840906A
- Authority
- CN
- China
- Prior art keywords
- physical
- lithofacies
- parameter
- reservoir
- property parameter
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000000704 physical effect Effects 0.000 title claims abstract description 68
- 238000011156 evaluation Methods 0.000 title claims abstract description 31
- 238000009826 distribution Methods 0.000 claims abstract description 61
- 238000005070 sampling Methods 0.000 claims abstract description 23
- 238000007619 statistical method Methods 0.000 claims abstract description 23
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 22
- 238000004088 simulation Methods 0.000 claims abstract description 21
- 239000004927 clay Substances 0.000 claims abstract description 18
- 230000008859 change Effects 0.000 claims abstract description 14
- 239000011435 rock Substances 0.000 claims abstract description 14
- 238000000034 method Methods 0.000 claims description 36
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 12
- 238000000342 Monte Carlo simulation Methods 0.000 claims description 11
- 238000005295 random walk Methods 0.000 claims description 9
- 238000013179 statistical model Methods 0.000 claims description 5
- 208000035126 Facies Diseases 0.000 claims description 3
- 230000000875 corresponding effect Effects 0.000 claims 2
- 238000011161 development Methods 0.000 abstract description 7
- 238000012502 risk assessment Methods 0.000 abstract description 3
- 230000008569 process Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000005553 drilling Methods 0.000 description 2
- 239000012530 fluid Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 238000013076 uncertainty analysis Methods 0.000 description 2
- 239000004215 Carbon black (E152) Substances 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 229930195733 hydrocarbon Natural products 0.000 description 1
- 150000002430 hydrocarbons Chemical class 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 229910052500 inorganic mineral Inorganic materials 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000011707 mineral Substances 0.000 description 1
- 238000011158 quantitative evaluation Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 230000000087 stabilizing effect Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 238000009827 uniform distribution Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V1/00—Seismology; Seismic or acoustic prospecting or detecting
- G01V1/28—Processing seismic data, e.g. analysis, for interpretation, for correction
- G01V1/30—Analysis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V1/00—Seismology; Seismic or acoustic prospecting or detecting
- G01V1/28—Processing seismic data, e.g. analysis, for interpretation, for correction
- G01V1/30—Analysis
- G01V1/306—Analysis for determining physical properties of the subsurface, e.g. impedance, porosity or attenuation profiles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
- G01V2210/00—Details of seismic processing or analysis
- G01V2210/60—Analysis
- G01V2210/62—Physical property of subsurface
- G01V2210/624—Reservoir parameters
- G01V2210/6244—Porosity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/08—Probabilistic or stochastic CAD
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
- Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping
Abstract
The invention provides an uncertainty evaluation method for reservoir physical property parameter prediction, which comprises the following steps: step 1: establishing a statistical petrophysical model of the target area, and deducing the conditional probability of the seismic attribute; step 2: carrying out statistical analysis on the existing physical parameter data, and establishing corresponding physical parameter prior distribution; step 3: based on the statistical petrophysical model and the prior probability distribution obtained in the step 1 and the step 2, carrying out lithofacies prediction by adopting a lithofacies variation physical parameter simulation sampling algorithm and generating a physical parameter sample set of a target area; step 4: and (3) obtaining a physical property parameter sample set according to the step (3), and estimating the expected, variance and confidence interval of the physical property parameters. The uncertainty evaluation method for reservoir physical property parameter prediction realizes statistical expectation, variance and confidence interval estimation of the clay content, porosity and saturation in the reservoir with rock phase change, so as to evaluate the potential risk of target reservoir development and provide technical support for risk analysis of oil and gas exploration development.
Description
Technical Field
The invention relates to the technical field of oilfield development, in particular to an uncertainty evaluation method for reservoir physical property parameter prediction.
Background
Seismic methods are one of the most effective technological means in the field of oil and gas exploration. Researchers can extract key attributes and important parameters from the seismic data that effectively reflect subsurface conditions. Among these, the physical parameters (shale content, porosity and saturation) are key to describing reservoir properties. In practice, the estimation of reservoir physical parameters directly or indirectly with seismic data is always subject to uncertainty due to unavoidable random factors such as noise in the data acquisition, assumed bias in the seismic interpretation, etc. In order to quantitatively express the uncertainty, a probabilistic model of physical parameters of a reservoir is established by combining a petrophysical model and a statistical analysis method, a matched simulation sampling algorithm is developed, and a brand new physical parameter prediction and uncertainty evaluation technology is developed.
One type of method in the field of seismic-based reservoir descriptions is a data-driven method. Such methods are often based on multivariate statistical methods that rely on well log data to build probability distributions of reservoir key properties (Doyen 1988;Fournier 1989). Due to the lack of reliable physical explanation and single statistical model structure, such methods have difficulty in solving the multi-parameter combination problem of the clay content, porosity and saturation. Another class of methods is model driven, building an empirical model based on petrophysical theory or laboratory data, and further achieving multiparameter joint prediction of clay content, porosity and saturation through inversion of the model (Nie et al 2004; fang and Yang 2015). However, such methods are deterministic in nature and do not take into account random factors in the process, and cannot evaluate uncertainties in predictions.
Both of the above approaches have met with some degree of success, however, a more potential approach is to combine the advantages of both. In 2006, bachrach describes randomness of seismic attribute data and logging data by introducing normal distribution and uniform distribution on the basis of a common petrophysical model (Gassmann model), joint estimation of porosity and gas saturation is realized, and variance of parameters is analyzed, but the method only aims at single sandstone, and influence of rock phase change is ignored. Since 2010 Grana et al set up different statistical petrophysical models based on a Bayesian statistical framework, and achieved the predictions of clay content, porosity and saturation and their uncertainty analysis (Grana and Rossa 2010; fig. 2017;Grana 2018;Li et al 2020). However, these methods all estimate the physical parameters directly based on the seismic data, and the forward process is computationally expensive and the inversion process has a strong initial dependency. Therefore, the posterior probability simulation sampling process of the physical property parameters often requires a large amount of calculation time, and it is difficult to traverse the probability support set, which results in low uncertainty evaluation efficiency and inaccuracy.
In short, in order to accurately and efficiently predict physical parameters and evaluate uncertainty of the physical parameters for a reservoir with a rock phase change, a complex petrophysical model needs to be built based on seismic attribute data, and a physical parameter sampling algorithm for the rock phase change is developed.
In application number: in the chinese patent application CN201510392504.7, a quantitative prediction method for tight sandstone gas reservoir is related to, comprising the following steps: establishing a geological model of actual geological features of a research area; the inversion parameter, impedance probability distribution and variation function of pre-stack geostatistics are obtained through the analysis of drilling in a work area, and the geologic body and the probability body are obtained through the calculation of a Markov chain Monte Carlo method by combining lithology data, logging data and seismic data on the basis of a geologic model. According to the method, drilling shear wave data are obtained by establishing a geological model and a petrophysical model of a sandstone gas reservoir, and earthquake sensitive parameters of the gas reservoir are obtained by analyzing characteristics of the gas reservoir. The inversion technology of the invention is used for obtaining a specific geologic body and a probability body, thereby solving the technical problems of low longitudinal resolution of earthquake, superposition of gas reservoir and surrounding rock impedance and low prediction precision of a tight sandstone gas reservoir; the invention reduces the polynomials of seismic reservoir predictions.
In application number: in the chinese patent application CN201610707298.9, a shale gas TOC pre-stack seismic inversion prediction method is related, which comprises the following steps: step one, establishing a TOC inversion objective function of a shale reservoir; step two, shale TOC pre-stack inversion based on elastic impedance: establishing prior distribution of the reservoir TOC according to statistical analysis of logging data, randomly sampling the established prior distribution through Monte Carlo simulation technology, finally obtaining random sample space distribution of the reservoir TOC, estimating the maximum value of posterior probability of the reservoir TOC, and obtaining the final inversion result by the TOC value corresponding to the position of the maximum value. The invention comprehensively applies the theories of Bayes theory, statistical rock physical model, monte Carlo random sampling technology and the like, can invert several physical parameters at the same time, eliminates the influence of other parameter limitations when inverting one parameter independently, and further enhances the reliability of inversion.
In application number: in CN201911138179.6, a method for predicting physical parameters of a reservoir layer by combining deep learning is related to the following steps: introducing nonlinear correlation of MIC quantitative measurement physical parameters and logging curves, and selecting logging curves with obvious response to the physical parameters; introducing CEEMDAN to decompose the physical property parameter data sequence to obtain an intrinsic mode function IMF component and a residual RES component, and stabilizing the physical property parameter data sequence; introducing complexity evaluation of SE on each IMF component and RES allowance, and recombining component sequences with similar entropy values to obtain new eigen mode components; the new intrinsic mode component data is divided into a training set and a testing set after being normalized; introducing an LSTM cyclic neural network to establish a prediction model for the reconstructed new component, and obtaining a predicted value of each new intrinsic mode component; and (3) inversely normalizing the predicted value of each new eigenmode component and carrying out superposition reconstruction to obtain a physical parameter predicted result. The method reduces the building modulus of redundant information and predicted components, and improves the prediction precision and the prediction speed.
The prior art is greatly different from the method, the technical problem to be solved by the method is not solved, and a novel uncertainty evaluation method for reservoir physical property parameter prediction is invented.
Disclosure of Invention
The invention aims to provide an uncertainty evaluation method for reservoir physical property parameter prediction, which is used for realizing statistical expectation, variance and confidence interval estimation of the clay content, porosity and saturation in a reservoir with lithofacies change so as to evaluate potential risks of target reservoir development.
The aim of the invention can be achieved by the following technical measures: an uncertainty evaluation method for reservoir physical property parameter prediction, the uncertainty evaluation method for reservoir physical property parameter prediction comprising:
step 1: establishing a statistical petrophysical model of the target area, and deducing the conditional probability of the seismic attribute;
step 2: carrying out statistical analysis on the existing physical parameter data, and establishing corresponding physical parameter prior distribution;
step 3: based on the statistical petrophysical model and the prior probability distribution obtained in the step 1 and the step 2, carrying out lithofacies prediction by adopting a lithofacies variation physical parameter simulation sampling algorithm and generating a physical parameter sample set of a target area;
step 4: and (3) obtaining a physical property parameter sample set according to the step (3), and estimating the expected, variance and confidence interval of the physical property parameters.
The aim of the invention can be achieved by the following technical measures:
in step 1, the statistical petrophysical model is composed of deterministic petrophysical model and random factors obtained by statistical analysis of logging data, is a model-data driven statistical model, and the conditional probability of the derived seismic attribute is easy to calculate.
In the step 2, the prior distribution is obtained by utilizing a classification or clustering algorithm to carry out statistical analysis on logging data, and comprises the prior probability of lithofacies and the prior distribution of physical parameters of fixed lithofacies; wherein the physical property parameter prior distribution of the fixed lithology is single-mode and easy to calculate and sample.
In step 3, the physical property parameter simulation sampling algorithm of the lithofacies change comprises probability estimation based on a Monte Carlo method and complex probability density simulation based on a Markov chain.
In step 3, based on the conditional expectation of the physical property parameters, the conditional probability of the lithofacies is estimated using the Monte Carlo method.
In step 3, aiming at the lithofacies with the highest probability, constructing a Markov chain of uniform random walk based on physical property parameter prior distribution of the fixed lithofacies and conditional probability of seismic attribute; the stationary distribution of the Markov chain is posterior distribution of physical parameters, and further, sufficient physical parameter samples are generated.
In step 3, for writing convenience, the statistical parameters are reservoir physical parameters including the clay content c, the porosity phi and the water saturation s, and are marked as l= (c, phi, s) ∈ [ L ] l ,U l ]，U l And L l Respectively representing the upper and lower boundaries of physical parameters in the physical sense; selecting specific seismic attribute data as observation data, and marking the data as d; the lithology is denoted f.
In step 3, according to the Bayesian statistical method, the posterior distribution p (l|d, f) of the physical property parameters of the stationary lithology is proportional to the product of the prior distribution p (l|f) of the corresponding physical property parameters and the conditional distribution p (d|l, f) of the seismic properties, i.e.
p(l|d,f)∝p(d|l,f)p(l|f). (3)
The posterior probability p (f|d) of a facies may represent a conditionally desired form of l|f, i.e
p(f|d)∝p(f)E l|f [p(d|l,f)] (4)
Wherein E is l|f Indicating the conditional expectation of l|f.
Calculating posterior probability of the lithofacies by a Monte Carlo method according to a formula (6) to realize prediction of the target lithofacies; and (3) constructing posterior distribution of physical property parameters under the random walk Markov chain simulation target lithofacies according to the formula (5).
In step 3, the physical property parameter simulation sampling algorithm of the lithofacies change specifically comprises:
first, for the k-th lithofacies f k According to the distribution p (l|f k ) Generating N k Sample { l } of physical parameters i Probability of lithofacies condition is calculated
Second, P is selected fk At maximum, correspond to lithofacies f M ；
Third, let the target distribution F (l) =p (d|l, F M )p(l|f M ) Step size delta of random walk, from initial distribution p (l|f M ) Is randomly generated (t) The current instant t=0 of the markov chain;
fourth, generating candidate value l * ＝l (t) +s,s～Unif(-δ - (l (t) ),δ + (l (t) ) And) wherein
x＝l (t) ，Corrected travel step length representing clay content, porosity and water-bearing wave saturation, respectively
Fifth step, calculating candidate value probability p t ＝min(R(l (t) ,l * ) 1), wherein
Sixth, generating a physical parameter sample l (t+1) ＝ql * +(1-q)l (t) Wherein q obeys the parameter p t 0-1 distribution of (2);
seventh, returning to the fourth step at the current time t=t+1 of the markov chain; until the Markov chain reaches a stop time T, namely T is more than T, stopping the operation, and outputting a physical property parameter sample set S l ＝{l (0) ,l (1) ,...,l (T) }；
Eighth step, based on the physical property parameter sample set S l Calculating a sample mean value to predict physical parameters, calculating a sample variance to quantitatively evaluate the risk of physical parameter prediction, and calculating a sample high-low score number to represent the possible variation range of the physical parameters.
In step 4, let T 0 Expected estimation of the clay content, indicative of burn-in periodDesired estimation of porosity->And a desired estimate of the saturation of water +.>Respectively is
According to the uncertainty evaluation method for reservoir physical parameter prediction, a deterministic petrophysical model and random errors are utilized to describe seismic attribute changes, a Bayesian statistical method is adopted to introduce prior distribution to describe randomness of physical parameters, a statistical petrophysical model is constructed, posterior distribution of the physical parameters is further analyzed, an efficient mixed probability model sampling algorithm is developed by combining a Markov chain and a Monte Carlo algorithm, and corresponding statistical feature quantity is estimated finally based on the generated clay content, porosity and saturation sample set.
The method fully considers the influence of random factors on the prediction of physical parameters, and is closer to the actual exploration condition. Test results show that in the aspect of oil and gas exploration, the method can simultaneously realize three tasks of lithology recognition, physical property parameter prediction and uncertainty evaluation, and has strong reservoir recognition capability and fluid recognition capability. The uncertainty evaluation method for reservoir physical property parameter prediction realizes statistical expectation, variance and confidence interval estimation of the clay content, porosity and saturation in the reservoir with rock phase change, and further evaluates potential risks of target reservoir development. Based on the rock physical modeling and statistical sampling ideas, a brand new physical parameter prediction and uncertainty evaluation thereof are developed aiming at reservoirs with rock phase changes, and further technical support is provided for risk analysis of oil and gas exploration and development.
Drawings
FIG. 1 is a schematic diagram of a physical property parameter simulation sampling algorithm in accordance with an embodiment of the present invention;
FIG. 2 is a schematic representation of seismic attribute data in an embodiment of the invention;
FIG. 3 is a diagram of lithofacies prediction results according to an embodiment of the present invention;
FIG. 4 is a diagram illustrating the results of physical property parameter estimation and confidence interval estimation according to an embodiment of the present invention;
FIG. 5 is a diagram illustrating estimation of variance of physical properties according to an embodiment of the present invention;
FIG. 6 is a flow chart of an embodiment of a method for uncertainty evaluation of reservoir property parameter predictions of the present invention.
Detailed Description
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the invention. Unless defined otherwise, 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 belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the present invention. As used herein, the singular forms also are intended to include the plural forms unless the context clearly indicates otherwise, and furthermore, it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, and/or combinations thereof.
The uncertainty evaluation method for predicting reservoir physical parameters (clay content, porosity and saturation) comprises the following steps: based on the related geological data and logging data of the target area, combining with a petrophysical theory, establishing a statistical petrophysical model of the target area; based on the statistical analysis of the existing physical property parameter data, establishing a priori model of the target area; based on the statistical petrophysical model and the prior model, the posterior probability simulation sampling of lithofacies and physical parameters is carried out, so that a physical parameter sample set is generated; according to the sample set, a statistical analysis method is adopted to realize the prediction of reservoir physical parameters of a target area and the uncertainty evaluation of the reservoir physical parameters. Based on the rock physical modeling and statistical sampling ideas, the method develops a brand new physical parameter prediction and uncertainty evaluation aiming at reservoirs with rock phase changes, and further provides technical support for risk analysis of oil and gas exploration and development.
The following are several embodiments of the invention
Example 1
In a specific embodiment 1 to which the present invention is applied, as shown in fig. 6, the uncertainty evaluation method of reservoir physical property parameter prediction of the present invention includes the steps of:
step 1: according to the existing various data, a statistical petrophysical model of the target area is established, and therefore the conditional probability of the seismic attribute can be deduced.
The statistical petrophysical model is composed of deterministic petrophysical model and random factors obtained by statistical analysis of logging data, is a model-data driven statistical model, and the conditional probability of the derived seismic attribute is convenient to calculate.
Step 2: and carrying out statistical analysis on the existing physical property parameter data, and establishing corresponding physical property parameter prior distribution. The prior distribution is obtained by using a classification (or clustering) algorithm to carry out statistical analysis on logging data, and comprises the prior probability of lithofacies and the prior distribution of physical parameters of fixed lithofacies. Wherein the physical property parameter prior distribution of the fixed lithology is single-mode and easy to calculate and sample.
Step 3: based on the step 1 and the step 2, a statistical petrophysical model and prior probability distribution are obtained, and a physical property parameter simulation sampling algorithm of lithofacies change is adopted to realize lithofacies prediction and generate a physical property parameter sample set of a target area.
The physical property parameter simulation sampling algorithm of the lithofacies change comprises probability estimation based on a Monte Carlo method and complex probability density simulation based on a Markov chain.
Based on the condition expectation of physical parameters, a Monte Carlo method is adopted to estimate the condition probability of lithofacies.
Aiming at the lithofacies with the highest probability, a Markov chain with uniform random walk is constructed based on the physical parameter prior distribution of the fixed lithofacies and the conditional probability of the seismic attribute. The stationary distribution of the Markov chain is posterior distribution of physical parameters, and further, sufficient physical parameter samples are generated.
Step 4: and (3) obtaining a physical property parameter sample set according to the step (3), and estimating the expected, variance and confidence interval of the physical property parameters.
Example 2
In a specific embodiment 2 to which the present invention is applied, in order to solve the problems existing in the prior art, the theoretical method adopted in the present invention mainly includes: bayesian statistical methods, monte carlo algorithms and MCMC theory.
For writing, the statistical parameters are reservoir physical parameters (clay content c, porosity phi and water saturation s) and are recorded as l= (c, phi, s) ∈ [ L ] l ,U l ]The method comprises the steps of carrying out a first treatment on the surface of the Selecting specific seismic attribute data as observation data, and marking the data as d; the lithology is denoted f.
Bayesian statistical methods have been widely used in the field of hydrocarbon prediction and reservoir description. In the invention, the Bayesian statistical method effectively links the rock phase identification and the physical property parameter prediction, and inhibits potential interference of rock phase change in the physical property parameter probability simulation sampling process. According to Bayesian statistical methods, the posterior distribution p (l|d, f) of a physical property parameter of a stationary lithology is proportional to the product of the prior distribution p (l|f) of the corresponding physical property parameter and the conditional distribution p (d|l, f) of the seismic property, i.e.
p(l|d,f)∝p(d|l,f)p(l|f). (5)
The posterior probability p (f|d) of a facies may represent a conditionally desired form of l|f, i.e
p(f|d)∝p(f)E l|f [p(d|l,f)] (6)
And (3) calculating posterior probability of the lithofacies by a Monte Carlo method according to the formula (6) to realize the prediction of the target lithofacies. Further, we construct a random walk Markov chain modeling the posterior distribution of physical parameters under the target lithology according to equation (5). As shown in fig. 1, a specific algorithm flow of the physical property parameter simulation sampling algorithm of the lithofacies change is as follows:
first, for any lithofacies f k According to the distribution p (l|f k ) Generating N k Sample { l } of physical parameters i Probability of lithofacies condition is calculated
Second, P is selected fk At maximum, correspond to lithofacies f M ；
Third, let the target distribution F (l) =p (d|l, F M )p(l|f M ) Step size delta of random walk, from initial distribution p (l|f M ) Is randomly generated (t) ,t＝0；
Fourth, generating candidate value l * ＝l (t) +s,s～Unif(-δ - (l (t) ),δ + (l (t) ) And) wherein
Fifth step, calculating candidate value probability p t ＝min(R(l (t) ,l * ) 1), wherein
Sixth, generating a physical parameter sample l (t+1) ＝ql * +(1-q)l (t) Wherein q obeys the parameter p t 0-1 distribution of (2);
seventh, t=t+1, returning to the fourth step; until enough physical parameter samples are generated, namely T is more than T, stopping operation, and outputting a physical parameter sample set S l ＝{l (0) ,l (1) ,...,l (T) }。
Eighth step, based on the physical property parameter sample set S l The method comprises the steps of calculating a sample mean value to predict physical parameters, calculating a sample variance to quantitatively evaluate the risk of physical parameter prediction, and calculating the number of high-low scores of the sample to represent the possible variation range of the physical parameters.
Example 3
In a specific embodiment 3 of the present invention, the present invention provides an uncertainty evaluation method for reservoir physical property parameter prediction, based on reservoir parameter prediction, by simulating and sampling the conditional distribution of physical property parameters, the present invention utilizes common seismic inversion attribute data (density, P-wave impedance and S-wave impedance) to estimate the reservoir clay content, porosity and water saturation, and gives corresponding uncertainty analysis results.
First, deterministic petrophysical models need to be built with reference to the properties of the common minerals and fluids of the target area and the associated geological data. Furthermore, the fitting error of the deterministic model is statistically analyzed according to the existing logging data, and a statistical petrophysical model is established, so that the conditional probability p (d|l, f) of the seismic attribute can be deduced. Meanwhile, a classification (or clustering) algorithm is adopted to carry out statistical analysis on the existing physical parameter data, the lithofacies category and the prior probability p (f) thereof are determined, and then the corresponding physical parameter prior distribution p (l|f) is established. Note that we should choose the appropriate seismic attribute and build an efficient petrophysical model to ensure that the conditional probability of the seismic attribute p (d|l, f) is computationally convenient; when the prior model is established, a proper statistical model is selected to ensure that the prior probability p (l|f) of the physical parameters is single-mode and easy to calculate and sample.
And secondly, based on the statistical petrophysical model and the prior probability distribution, adopting a physical property parameter simulation sampling algorithm of the lithofacies change shown in figure 1 to realize lithofacies prediction and generate a physical property parameter sample set of the target area. Further, the expected, variance and confidence interval of the physical property parameters are estimated from the physical property parameter sample set.
FIG. 2 is seismic inversion attribute data for a test patent method. The data are poisson' S ratio, P-wave impedance and S-wave impedance simulated from known physical property parameter logging data. Wherein, the black line is deterministic petrophysical model synthetic data, and the gray line is statistical petrophysical model generating data.
Fig. 3 is a lithofacies prediction result, the left graph is a real situation of lithofacies, and the right graph is a result of lithofacies prediction by the method of the patent. Wherein gray represents sandstone and black represents mudstone. Comparing the two figures, the sand bodies of 2595-2600ms and 2625-2630ms are correctly identified, which shows that the method can effectively identify sandstone.
Fig. 4 (a) - (c) are expected and confidence interval estimates for the clay content, porosity and water saturation, respectively. Wherein, the solid black line is the actual physical parameter, and the dashed black line is the physical parameter expected estimation. As can be seen from the figures, the solid black line is substantially identical to the dashed black line, but does not coincide exactly at the details. This suggests that there is always uncertainty in the quantitative predictions of physical parameters, but that a simple estimate of expectations cannot characterize such uncertainty. The grey shaded portion of the graph shows the 95% confidence interval for the physical property parameter, and we find that the black solid line falls completely within the grey shaded portion and that the width of the shaded portion is different. This means that the method can accurately estimate the confidence interval of the physical parameter, further give the random variation range of the physical parameter, and intuitively describe the uncertainty of the physical parameter.
Fig. 5 (a) - (c) are variance estimates of clay content, porosity and water saturation, respectively. As shown, the water saturation predictions for the sandstone zones at 2595-2600ms and 2625-2630ms are higher in variance, i.e., there is greater uncertainty in the water saturation predictions for the zones, which is consistent with qualitative insights in actual work. This shows that the method can effectively estimate the variance of the physical property parameters and further give quantitative evaluation indexes of uncertainty of the physical property parameter prediction.
Finally, it should be noted that: the foregoing description is only a preferred embodiment of the present invention, and is not intended to limit the present invention, but although the present invention has been described in detail with reference to the foregoing embodiment, it will be apparent to those skilled in the art that modifications may be made to the technical solution described in the foregoing embodiment, or equivalents may be substituted for some of the technical features thereof. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Other than the technical features described in the specification, all are known to those skilled in the art.
Claims (10)
1. The uncertainty evaluation method for reservoir physical property parameter prediction is characterized by comprising the following steps:
step 1: establishing a statistical petrophysical model of the target area, and deducing the conditional probability of the seismic attribute;
step 2: carrying out statistical analysis on the existing physical parameter data, and establishing corresponding physical parameter prior distribution;
step 3: based on the statistical petrophysical model and the prior probability distribution obtained in the step 1 and the step 2, carrying out lithofacies prediction by adopting a lithofacies variation physical parameter simulation sampling algorithm and generating a physical parameter sample set of a target area;
step 4: and (3) obtaining a physical property parameter sample set according to the step (3), and estimating the expected, variance and confidence interval of the physical property parameters.
2. The uncertainty evaluation method of reservoir property parameter prediction according to claim 1, wherein in step 1, the statistical petrophysical model is composed of deterministic petrophysical model and random factors obtained by statistical analysis of well logging data, is a model-data driven statistical model, and the conditional probability of the derived seismic attribute is computationally convenient.
3. The uncertainty evaluation method of reservoir physical property parameter prediction according to claim 1, wherein in step 2, the prior distribution is obtained by performing statistical analysis on the logging data by using a classification or clustering algorithm, and includes a lithofacies prior probability and a physical property parameter prior distribution of a fixed lithofacies; wherein the physical property parameter prior distribution of the fixed lithology is single-mode and easy to calculate and sample.
4. The uncertainty evaluation method for reservoir property parameter prediction according to claim 1, wherein in step 3, the property parameter simulation sampling algorithm for rock phase change includes probability estimation based on a monte carlo method and complex probability density simulation based on a markov chain.
5. The uncertainty evaluation method for reservoir property parameter prediction as described in claim 4, wherein in step 3, based on the conditional expectation of the property parameter, the conditional probability of the lithofacies is estimated by using a monte carlo method.
6. The uncertainty evaluation method for reservoir physical property parameter prediction according to claim 5, wherein in step 3, a uniform random walk markov chain is constructed for the lithofacies with the highest probability based on the physical property parameter prior distribution of the fixed lithofacies and the conditional probability of the seismic attribute; the stationary distribution of the Markov chain is posterior distribution of physical parameters, and further, sufficient physical parameter samples are generated.
7. The reservoir physical property of claim 6The uncertainty evaluation method of parameter prediction is characterized in that in step 3, for writing convenience, the statistical parameters are reservoir physical parameters including clay content c, porosity phi and water saturation s, and are marked as l= (c, phi, s) ∈ [ L ] l ,U l ]，U l And L l Respectively representing the upper and lower boundaries of physical parameters in the physical sense; selecting specific seismic attribute data as observation data, and marking the data as d; the lithology is denoted f.
8. The uncertainty evaluation method of reservoir property parameter prediction as claimed in claim 7, wherein in step 3, the property parameter posterior distribution p (l|d, f) of the fixed lithology is proportional to the product of the prior distribution p (l|f) of the corresponding property parameter and the conditional distribution p (d|l, f) of the seismic attribute, i.e., according to bayesian statistical method
p(l|d,f)∝p(d|l,f)p(l|f) (1)
The posterior probability p (f|d) of a facies may represent a conditionally desired form of l|f, i.e
p(f|d)∝p(f)E l|f [p(d|l,f)] (2)
Wherein E is l|f A conditional expectation representing l|f;
calculating posterior probability of the lithofacies by a Monte Carlo method according to a formula (6) to realize prediction of the target lithofacies; and (3) constructing posterior distribution of physical property parameters under the random walk Markov chain simulation target lithofacies according to the formula (5).
9. The uncertainty evaluation method for reservoir property parameter prediction as described in claim 8, wherein in step 3, the property parameter simulation sampling algorithm for lithofacies changes specifically comprises:
first, for the k-th lithofacies f k According to the distribution p (l|f k ) Generating N k Sample { l } of physical parameters i Probability of lithofacies condition is calculated
In the second step, the second step is carried out,selectingAt maximum, correspond to lithofacies f M ；
Third, let the target distribution F (l) =p (d|l, F M )p(l|f M ) Step size delta of random walk, from initial distribution p (l|f M ) Is randomly generated (t) ＝(c (t) ,φ (t) ,s (t) ) The current time t=0 of the markov chain;
fourth, generating candidate value l * ＝l (t) +s,s～Unif(-δ - (l (t) ),δ + (l (t) ) And) wherein
x＝l (t) ，The corrected walk-away steps respectively representing the clay content, the porosity and the saturation of the water-containing wave;
fifth step, calculating candidate value probability p t ＝min(R(l (t) ,l * ) 1), wherein
Sixth, generating a physical parameter sample l (t+1) ＝ql * +(1-q)l (t) Wherein q obeys the parameter p t 0-1 distribution of (2);
seventh, returning to the fourth step at the current time t=t+1 of the markov chain; until the Markov chain reaches the stop time T, i.e. T is more than T, stopping the operation and outputting objectsSexual parameter sample set S l ＝{l (0) ,l (1) ,...,l (T) }；
Eighth step, based on the physical property parameter sample set S l Calculating a sample mean value to predict physical parameters, calculating a sample variance to quantitatively evaluate the risk of physical parameter prediction, and calculating a sample high-low score number to represent the possible variation range of the physical parameters.
10. The method for evaluating uncertainty of reservoir property parameter prediction according to claim 1, wherein in step 4, T is set up 0 Expected estimation of the clay content, indicative of burn-in periodDesired estimation of porosity->And a desired estimate of the saturation of water +.>Respectively is
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210288063.6A CN116840906A (en) | 2022-03-23 | 2022-03-23 | Uncertainty evaluation method for reservoir physical property parameter prediction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210288063.6A CN116840906A (en) | 2022-03-23 | 2022-03-23 | Uncertainty evaluation method for reservoir physical property parameter prediction |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116840906A true CN116840906A (en) | 2023-10-03 |
Family
ID=88160344
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210288063.6A Pending CN116840906A (en) | 2022-03-23 | 2022-03-23 | Uncertainty evaluation method for reservoir physical property parameter prediction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116840906A (en) |
-
2022
- 2022-03-23 CN CN202210288063.6A patent/CN116840906A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11668853B2 (en) | Petrophysical inversion with machine learning-based geologic priors | |
| US10901105B2 (en) | Method and system for regression and classification in subsurface models to support decision making for hydrocarbon operations | |
| EP3894902B1 (en) | Subsurface models with uncertainty quantification | |
| Syed et al. | Application of ML & AI to model petrophysical and geomechanical properties of shale reservoirs–A systematic literature review | |
| US20200124753A1 (en) | Method and System for Evaluating Variability in Subsurface Models to Support Decision Making for Hydrocarbon Operations | |
| US10768324B2 (en) | Method to predict pore pressure and seal integrity using full wavefield inversion | |
| US8255165B2 (en) | Method for predicting differences in subsurface conditions | |
| Grana et al. | Probabilistic inversion of seismic data for reservoir petrophysical characterization: Review and examples | |
| Iturrarán-Viveros | Smooth regression to estimate effective porosity using seismic attributes | |
| Fu et al. | A blocking Markov chain Monte Carlo method for inverse stochastic hydrogeological modeling | |
| Shi et al. | Brittleness index prediction in shale gas reservoirs based on efficient network models | |
| US8965744B2 (en) | Method of developing a petroleum reservoir by reservoir model reconstruction | |
| Mosser et al. | Deepflow: history matching in the space of deep generative models | |
| You et al. | Shear wave velocity prediction based on LSTM and its application for morphology identification and saturation inversion of gas hydrate | |
| Cornelio et al. | Physics-assisted transfer learning for production prediction in unconventional reservoirs | |
| WO2020142257A1 (en) | Method and system for evaluating variability in subsurface models to support decision making for hydrocarbon operations | |
| Jeong et al. | Application of conditional generative model for sonic log estimation considering measurement uncertainty | |
| Xie et al. | Reservoir facies design and modeling using probabilistic rock-physics templates | |
| Booth et al. | Grid-based inversion of pressure transient test data with stochastic gradient techniques | |
| Lu et al. | Improved estimation and forecasting through residual-based model error quantification | |
| Bazargan | An efficient polynomial chaos-based proxy model for history matching and uncertainty quantification of complex geological structures | |
| CN116840906A (en) | Uncertainty evaluation method for reservoir physical property parameter prediction | |
| Li | Machine learning for the subsurface characterization at core, well, and reservoir scales | |
| Choubineh et al. | Machine learning to improve natural gas reservoir simulations | |
| Osman et al. | Machine Learning Approach to Rock Properties Determination using NMR Log Data. |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |