A method for estimating formation pore pressure while drilling based on Bayesian theory

Abstract

The present invention discloses a method for estimating formation pore pressure while drilling based on Bayesian theory, comprising the following steps: collecting target well logging data, while drilling logging data, and data from adjacent wells in the same block as the target well; predicting the target well's confidence-containing formation pressure interval profile based on the logging data and while drilling logging data, and determining the prior probability of formation pressure at any well depth; calculating the confidence-containing formation pressure interval profile of the adjacent well based on the adjacent well data, and determining the likelihood function of formation pressure at any well depth; selecting the prior probability and likelihood function of formation pressure at the same layer and substituting them into the Bayesian formula to obtain an updated posterior probability of formation pressure. The present invention realizes the while drilling update and correction of the confidence-containing formation pressure interval profile, thereby more accurately obtaining formation pressure parameters; providing more accurate formation pressure information for drilling risk assessment, reducing drilling risks caused by unclear understanding of formation pressure, and effectively improving drilling efficiency.

Description

Technical Field

[0001] The invention belongs to the technical field of complex formation drilling petroleum, and in particular relates to a formation pore pressure while drilling estimation method based on Bayesian theory. Background Art

[0002] Formation pressure, also known as formation pore pressure, refers to the pressure acting on the fluid within the rock pores. It reflects fundamental formation parameters such as the fluid conditions, rock type, rock mechanical properties, and geological structure. Formation pressure information is essential for ensuring safe, efficient, and economical drilling operations. Accurately estimating formation pressure helps drillers select the appropriate drilling fluid density, prevent safety incidents such as kicks and blowouts caused by abnormally high pressure, and reduce drilling time. Furthermore, estimated pressure information facilitates the design of appropriate wellbore structures, reducing construction costs while ensuring safety. Therefore, formation pressure estimation is crucial in oil and gas resource exploration, oil and gas field development, and drilling engineering.

[0003] Currently, there are many methods for estimating formation pressure. These methods, based on the drilling sequence, can be categorized as follows: pre-drilling pressure prediction, pressure monitoring while drilling (LWD), post-drilling log interpretation, and actual pressure measurement. Pre-drilling pressure prediction primarily utilizes seismic data. Seismic layer velocities are obtained through data processing, and a theoretical relationship model between seismic layer velocities and formation pressure is established to calculate formation pressure. Common methods include the Fillippone method and the single-point prediction method. Pressure monitoring while drilling (LWD) utilizes LWD logging data and related engineering data to monitor formation pressure in real time. Common methods include the DC index, σ index, and standardized rate of penetration. Post-drilling log interpretation is the most commonly used and generally considered the most accurate method. Common methods include the equivalent depth method, the Eaton method, and the effective stress method. Actual pressure measurement involves direct instrumentation to measure formation pressure. Common methods include drill stem testing and repeated formation testing.

[0004] However, due to the complex spatial variability of geological bodies, different understandings of the causes of pressure can lead to different theoretical estimation models, and the inherent inherent errors in prediction data (seismic, mud logging, and well logging data) lead to significant uncertainty in formation pressure estimates. When drilling operations rely on such uncertain formation pressure information to formulate plans, deviations in drilling operations can occur, and even serious drilling accidents can occur.

[0005] Therefore, in order to take into account the influence of various uncertainties on formation pressure estimation, the present invention aims to provide a formation pore pressure estimation method while drilling based on Bayesian theory. Summary of the Invention

[0006] The problem to be solved by the present invention is to provide a method for estimating formation pore pressure while drilling based on Bayesian theory. This method uses a formation pressure interval profile containing confidence levels and updates the target well formation pressure estimation result through adjacent well data, thereby improving the accuracy of formation pressure estimation and providing scientific and effective technical support for oil and gas field exploration and development and drilling risk assessment.

[0007] To solve the above technical problems, the present invention adopts a technical solution: a formation pore pressure estimation method while drilling based on Bayesian theory, comprising the following steps:

[0008] S1: Collecting well logging data, logging while drilling data of the target well and data of adjacent wells in the same block as the target well;

[0009] S2: predicting a confidence-based formation pressure interval profile of the target well based on the well logging data and the logging while drilling data, and determining a priori probability of formation pressure at any well depth;

[0010] S3: Based on the adjacent well data, calculate the confidence interval profile of the adjacent well and determine the likelihood function of the formation pressure at any well depth;

[0011] S4: Select the prior probability and likelihood function of the formation pressure at the same layer and substitute them into the Bayesian formula to obtain the updated posterior probability of the formation pressure.

[0012] Furthermore, the S2 includes the following steps:

[0013] S21: Calculating the formation pressure of the entire well section of the target well based on the effective stress method formation pressure prediction model;

[0014] S22: Substitute the formation pressure Pp of the entire well section calculated by the effective stress method into the DC index method formula, inversely calculate the Eaton index at any well depth along the vertical direction, and construct an analysis sample library of the Eaton index;

[0015] S23: Using the sliding window method, the Eaton index at any well depth is converted from a single sample into a set of 2n+1 small samples. The normal information diffusion method is used for probability distribution fitting to establish the probability density function and cumulative probability density function of the Eaton index at any well depth.

[0016] S24: Selecting a pure mudstone section as a normal compaction trend line construction interval, and defining a starting well section and an ending well section within the interval, wherein the starting well section contains n sample points, and the ending well section contains m sample points;

[0017] S25: Selecting the nth sample point in the starting well section and the mth sample point in the ending well section respectively, constructing n×m normal compaction trend lines, and constructing an analysis sample library of slopes and intercepts of the normal compaction trend lines;

[0018] S26: performing probability statistical analysis on the slope and intercept of the normal compaction trend line, selecting the normal information diffusion method for probability distribution fitting, and establishing a probability density function and a cumulative probability density function of the normal compaction trend line;

[0019] S27: Generate a random number sample XN that conforms to the probability density function of the characteristic parameters of the Eaton index and the normal compaction trend line through Monte Carlo simulation, where N is the number of Monte Carlo simulations;

[0020] S28: Substitute the random number sample generated in S27 into the DC index method calculation model to calculate the formation pressure sample set P at any well depth. prnd , select normal distribution to fit the formation pressure to obtain the prior probability density function f(Pp) and cumulative probability density function F(Pp) of the formation pressure;

[0021] S29: Take the pressure values ​​of the cumulative probability density of formation pressure J1=5% and J2=95% at any well depth, connect the points along the vertical well depth to form a line, and obtain the formation pressure curve with a cumulative probability of |J2-J1|=90% for the entire well section of the target well, that is, the formation pressure interval profile with a confidence level of 90%.

[0022] Furthermore, in S21, the effective stress method formation pressure prediction model is as follows:

[0023] P P =S v -100.674e -2.5782μ

[0024] Where: P P is the formation pressure, in MPa; S v is the overburden pressure, in MPa; μ is Poisson's ratio, dimensionless;

[0025] Calculate the overburden pressure S by using the logging data v And the Poisson's ratio μ, the formula is as follows:

[0026]

[0027] Where: ρ is the average density of the overlying rock layer, in g / cm 3 ρ b is the formation density, in g / cm 3 ; H0 is the starting depth of the target well, in m; H is the ending depth of the target well, in m; Δts is the shear wave time difference, unit is us / m; Δt P is the longitudinal wave time difference, unit is us / m.

[0028] Furthermore, in S22, the calculation model of the Eaton index is as follows:

[0029]

[0030] Where: n is the Eaton index; d c is the measured d c Index; d cn The normal compaction trend line of the formation corresponds to d c Index; P h is the hydrostatic pressure, in MPa; d c The index is calculated using drilling data using the following formula:

[0031]

[0032] Where: T is drilling time, unit is min / m; N is rotary table speed, unit is r / min; W is drilling pressure, unit is KN; D b is the drill diameter, in mm; G h is the hydrostatic pressure gradient, in g / cm 3 ρ ECD is the circulating equivalent drilling fluid density, in g / cm 3 .

[0033] Furthermore, in S23, the calculation model of the normal information diffusion method is as follows:

[0034]

[0035] Where: n is the sample size; h is the diffusion coefficient; x i is the observation sample;

[0036] When the sample volume n is different, the diffusion coefficient h is calculated by the following formula:

[0037]

[0038] Where: b is the maximum value of the sample, b=max{x i}; a is the minimum value of the sample, a=min{x i}; σ is the sample standard deviation; ζ is the correlation coefficient, which is determined by the sample size.

[0039] Furthermore, in said S28, the random sample set P of formation pressure at any well depth position prnd The calculation formula is as follows:

[0040]

[0041] Where: nrnd is the random number sample set of Eaton index; d cnrnd A sample set of random numbers for the normal compaction trend line.

[0042] Furthermore, in S28, the a priori probability density function and cumulative probability density function formulas of the target well formation pressure are as follows:

[0043]

[0044] Where: μ is the sample mean; σ is the sample standard deviation.

[0045] Furthermore, the S3 includes the following steps:

[0046] S31: According to the geological stratification of the block where the target well is located, screen and obtain information of neighboring wells in the same stratum as the target well;

[0047] S32: According to S2, the likelihood function L(X1|P) of the target layer formation pressure is obtained through the information of adjacent wells.

[0048] Furthermore, the S4 includes the following steps:

[0049] S41: The target well formation pressure prior probability f(P p ) and the likelihood function L(X 1 |P)Substitute into the Bayesian formula;

[0050] S42: Obtain the posterior probability g(P|X 1 ) to update the formation pressure of the target well.

[0051] Furthermore, the Bayesian formula is as follows:

[0052]

[0053] Where: p(X) is the marginal probability, that is, the normalization coefficient.

[0054] The advantages and positive effects of the present invention are:

[0055] This method combines the effective stress method with the DC index method to quantitatively characterize the uncertainty of the Eaton index and normal compaction trend line during formation pressure estimation through mathematical and statistical analysis. It also establishes a confidence-based formation pressure prediction interval profile for the target well. Furthermore, by considering information from neighboring wells in the same block and using Bayesian theory, it enables while-drilling updating and correction of this confidence-based formation pressure interval profile, thereby more accurately obtaining formation pressure parameters. This provides more accurate formation pressure information for drilling risk assessment, reduces drilling risks associated with unclear formation pressure understanding, and effectively improves drilling efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a schematic diagram of the overall process of an embodiment of the present invention.

[0057] Figure 2 2 is a schematic diagram of a sliding window method processing according to an embodiment of the present invention.

[0058] Figure 3 、 Figure 4 2 is a schematic diagram of a sample structure of characteristic parameters of a normal compaction trend line according to an embodiment of the present invention.

[0059] Figure 5 It is the formation pressure of the entire well section calculated by the effective stress method in the X-1 well in the specific embodiment of the present invention.

[0060] Figure 6 This is an Eaton index diagram at 3100 m in Well X-1 in a specific embodiment of the present invention.

[0061] Figure 7 、 Figure 8 This is a probability density distribution diagram of characteristic parameters of the normal compaction trend line at 3100 m in the X-1 well in a specific embodiment of the present invention.

[0062] Figure 9 This is a statistical graph of random number samples of the Eaton index at 3100 m in the X-1 well in a specific embodiment of the present invention.

[0063] Figure 10 、 Figure 11 This is a statistical chart of random number simulation samples of the normal compaction trend line at 3100 m in the X-1 well in a specific embodiment of the present invention.

[0064] Figure 12 It is the probability density distribution and cumulative probability density distribution diagram of the formation pressure at 3100m in the X-1 well in a specific embodiment of the present invention.

[0065] Figure 13 It is a cross-sectional diagram of the confidence formation pressure interval of the X-1 well in a specific embodiment of the present invention.

[0066] Figure 14It is a cross-sectional diagram of the confidence formation pressure interval of the X-2 well in a specific embodiment of the present invention.

[0067] Figure 15 It is a posterior probability distribution diagram of the formation pressure at 3100 m in a specific embodiment of the present invention.

[0068] Figure 16 It is a cross-sectional diagram of the formation pressure interval containing confidence after Bayesian updating in a specific embodiment of the present invention. DETAILED DESCRIPTION

[0069] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present invention.

[0070] The embodiments of the present invention are further described below with reference to the accompanying drawings:

[0071] like Figure 1 As shown, a method for estimating formation pore pressure while drilling based on Bayesian theory includes the following steps.

[0072] S1: Collect well logging data, logging while drilling data of the target well, and data of adjacent wells in the same block as the target well.

[0073] Specifically, well logging data includes sonic logging, density logging, natural gamma ray logging, and other well logging data. Logging while drilling data includes drilling time, rotary table speed, weight on bit, drill bit diameter, and circulating equivalent drilling fluid density of the target well.

[0074] S2: Based on the well logging data and the logging while drilling data, the target well's formation pressure interval profile with confidence is predicted, and the prior probability of the formation pressure at any well depth is determined. Specifically, S2 includes the following steps:

[0075] S21: Based on the collected logging and mud logging interpretation results of the target well, or by conducting acoustic wave velocity, uniaxial compression, and triaxial compression experiments, determine basic parameters of the target well, such as the P-wave time difference Δtp, S-wave time difference Δts, static Poisson's ratio μs, DC exponent dc, and overburden pressure Sv. Based on the effective stress method formation pressure prediction model, calculate the formation pressure for the entire target well section.

[0076] Specifically, the overburden pressure S v and Poisson's ratio μ, as follows:

[0077]

[0078] Where: ρ is the average density of the overlying rock layer, in g / cm 3 ρ b is the formation density, in g / cm 3 ; H0 is the starting depth of the target well, in m; H is the ending depth of the target well, in m; Δt s is the shear wave time difference, unit is us / m; Δt P is the longitudinal wave time difference, unit is us / m.

[0079] The formation pressure of the drilled formation is calculated using the effective stress method formation pressure prediction model:

[0080] P P =S v -100.674e -2.5782μ Formula (3)

[0081] Where: P P is the formation pressure, in MPa; S v is the overburden pressure, in MPa; μ is Poisson’s ratio, dimensionless.

[0082] S22: Substitute the formation pressure Pp of the entire well section calculated by the effective stress method into the DC index method formula, inversely calculate the Eaton index at any well depth along the vertical direction, and construct an analysis sample library of the Eaton index.

[0083] Specifically, the calculation model of the Eaton index is as follows:

[0084]

[0085] Where: n is the Eaton index; d c is the measured d c Index; d cn The normal compaction trend line of the formation corresponds to d c Index; P h is the hydrostatic pressure, in MPa; d c The index is calculated using drilling data using the following formula:

[0086]

[0087] Where: T is drilling time, unit is min / m; N is rotary table speed, unit is r / min; W is drilling pressure, unit is KN; D b is the drill diameter, in mm; G h is the hydrostatic pressure gradient, in g / cm 3 ρ ECD is the circulating equivalent drilling fluid density, in g / cm 3 .

[0088] S23: If Figure 2 As shown in the figure, the Eaton index at any well depth is converted from a single sample into a set of 2n+1 small samples through the sliding window method, and the normal information diffusion method is selected for probability distribution fitting to establish the probability density function and cumulative probability density function of the Eaton index at any well depth.

[0089] Specifically, the probability density function calculation formula is as follows:

[0090]

[0091] The cumulative probability density function calculation formula is as follows:

[0092]

[0093] Where: n is the sample size; h is the diffusion coefficient; x i is the observation sample.

[0094] When the sample volume n is different, the diffusion coefficient h is calculated as follows:

[0095]

[0096] Where: b is the maximum value of the sample, b=max{x i}; a is the minimum value of the sample, a=min{x i}; σ is the sample standard deviation; ζ is the correlation coefficient, which is determined by the sample size, as shown in Table 1.

[0097] Table 1 ζ values ​​corresponding to different sample sizes

[0098]

[0099] S24: Quantitatively characterize the uncertainty of the normal compaction trend line. Select the pure mudstone section as the normal compaction trend line construction interval, and define the starting well section and the ending well section within the interval. The starting well contains n sample points, and the ending well section contains m sample points, such as Figure 3 、 Figure 4 shown.

[0100] S25: Select the nth sample point in the starting well section and the mth sample point in the ending well section respectively, construct n×m normal compaction trend lines, and construct an analysis sample library of the slope and intercept of the normal compaction trend line.

[0101] S26: Perform probability statistical analysis on the slope and intercept of the normal compaction trend line, select the normal information diffusion method for probability distribution fitting, and establish the probability density function and cumulative probability density function of the normal compaction trend line.

[0102] S27: Generate a random number sample XN that conforms to the probability density functions of the characteristic parameters of the Eaton index and the normal compaction trend line through Monte Carlo simulation, where N is the number of Monte Carlo simulations.

[0103] S28: Substitute the random number sample generated in S27 into the DC index method calculation model to calculate the formation pressure sample set P at any well depth. prnd , the normal distribution is selected to fit the formation pressure to obtain the prior probability density function f(Pp) and cumulative probability density function F(Pp) of the formation pressure.

[0104] The random sample set P of formation pressure at any well depth prnd The calculation formula is as follows:

[0105]

[0106] Where: nrnd is the random number sample set of Eaton index; d cnrnd A sample set of random numbers for the normal compaction trend line.

[0107] In S28, the prior probability density function f(Pp) of the target well formation pressure is given by the following formula:

[0108]

[0109] The cumulative probability density function F(Pp) of the target well formation pressure is given by:

[0110]

[0111] Where: μ1 is the mean value of the target well formation pressure sample; σ1 is the standard deviation of the target well formation pressure sample.

[0112] S29: Take the pressure values ​​of the cumulative probability density of formation pressure J1=5% and J2=95% at any well depth, connect the points along the vertical well depth to form a line, and obtain the formation pressure curve with a cumulative probability of |J2-J1|=90% for the entire well section of the target well, that is, the formation pressure interval profile with a confidence level of 90%.

[0113] S3: Based on the data of adjacent wells, calculate the confidence interval profile of the adjacent wells and determine the likelihood function of the formation pressure at any well depth. Specifically, S3 includes the following steps:

[0114] S31: Based on the geological stratification of the block where the target well is located, the information of neighboring wells in the same layer as the target well is screened and obtained, including well logging data, mud logging data, etc.

[0115] S32: According to S2, the likelihood function L(X1|P) of the target layer formation pressure is obtained through the information of the adjacent wells, and a formation pressure interval profile with a confidence level of 90% for the adjacent wells is established.

[0116]

[0117] Where: μ2 is the mean value of the formation pressure samples of adjacent wells; σ2 is the standard deviation of the formation pressure samples of adjacent wells.

[0118] S4: Select the prior probability and likelihood function of the formation pressure at the same layer and substitute them into the Bayesian formula to obtain the updated posterior probability of the formation pressure. Specifically, S4 includes the following steps:

[0119] S41: The target well formation pressure prior probability f(P p ) and the likelihood function L(X 1 |P) into the Bayesian formula.

[0120] S42: Obtain the posterior probability g(P|X 1 ) to update the formation pressure of the target well.

[0121] The Bayesian formula is as follows:

[0122]

[0123] Where: p(X) is the marginal probability, that is, the normalization coefficient.

[0124] The present invention will be specifically described below using Well X-1 as an example:

[0125] S1: Collect acoustic logging, density logging, natural gamma ray and other logging data from Well X-1, as well as logging while drilling data such as drilling time, rotary table speed, bit weight, drill bit diameter and circulating equivalent drilling fluid density, as well as drilling data from the adjacent well X-2 in the same block.

[0126] S2: Based on the well logging data and the logging while drilling data, the prior probability of the formation pressure at any depth of well X-1 is calculated, and the formation pressure interval profile of X-1 with confidence is established.

[0127] S21: Calculate the formation pressure of the entire section of Well X-1 using the effective stress method, such as Figure 5 shown.

[0128] S22: Substitute the formation pressure Pp obtained by effective stress calculation of Well X-1 into the DC index calculation formula, inversely calculate the Eaton index, and obtain a sample analysis library of the Eaton index.

[0129] S23: The Eaton index is processed by the sliding window method to construct a small sample set of the Eaton index at any well depth. The probability density function f1(x) and the cumulative probability density function F1(x) of the Eaton index are constructed by the normal information diffusion method.

[0130] S24: The pure mudstone section of Well X-1 is selected as the structural interval of the normal compaction trend line.

[0131] S25: Constructing an analysis sample library of the slope and intercept of the normal compaction trend line according to the method mentioned in the present invention.

[0132] S26: The probability density functions f2(x), f3(x) and the cumulative probability density functions F2(x), F3(x) of the normal compaction trend line (slope and intercept) are constructed by the normal information diffusion method. The analysis is performed at the position of 3100m in the X-1 well, as shown in the following example: Figure 6-8 shown.

[0133] S27: Using MATLAB programming and Monte Carlo simulation method, generate random number samples XN that conform to the probability density functions of the characteristic parameters (slope and intercept) of the Eaton index and the normal compaction trend line, and set the number of simulations N = 104 times, as shown in Figure 9-11 shown.

[0134] S28: Substitute the generated random number samples of each parameter back into the DC index method (Formula 9) to obtain the formation pressure random sample set P prnd , select normal distribution to fit the formation pressure, and obtain the probability density function f(Pp) and cumulative probability density function F(Pp) of the formation pressure at 3100m, as shown in Figure 12 shown.

[0135] S29: Take the pressure values ​​of the formation pressure cumulative probability density J1 = 5% and J2 = 95% at any well depth, connect the points along the vertical well depth to form a line, and obtain the formation pressure curve with the cumulative probability of |J2-J1| = 90% for the entire well section of well X-1, that is, the formation pressure interval profile with a confidence level of 90%, as shown in Figure 13 shown.

[0136] S3: Based on the well logging data and logging while drilling data of the adjacent well X-2, the formation pressure interval profile with a confidence level of 90% is constructed in the same way as well X-2, such as Figure 14 shown.

[0137] S4: Select the formation pressure at 3100m of Well X-1 as the prior probability N(42.27,2.422) of the formation pressure at the depth of the well, select the formation pressure at 3100m of Well X-2 as the likelihood function N(54.76,3.062) of the formation pressure at the depth of the well, substitute the prior probability and likelihood function into the Bayesian formula, and calculate the posterior probability of the formation pressure at 3100m as N(46.87,1.062), as shown in the following example: Figure 15 shown.

[0138] Finally, the formation pressure information of well X-1 is used as the prior probability and the formation pressure information of well X-2 is used as the likelihood function. The formation pressure estimation results at any depth are updated using the Bayesian formula to obtain the updated confidence interval profile of the formation pressure based on the Bayesian theory, as shown in the following example: Figure 16 shown.

[0139] The updated posterior probability of formation pressure integrates the target well formation pressure prediction information and the adjacent well formation pressure prediction information. It is continuously revised and updated based on previous predictions using adjacent well data, ensuring the accuracy of local formation pressure estimation to the greatest extent possible and providing more reasonable formation pressure information for drilling operations.

[0140] The advantages and positive effects of the present invention are:

[0141] This method combines the effective stress method with the DC index method to quantitatively characterize the uncertainty of the Eaton index and normal compaction trend line during formation pressure estimation through mathematical and statistical analysis. It also establishes a confidence-based formation pressure prediction interval profile for the target well. Furthermore, by considering information from neighboring wells in the same block and using Bayesian theory, it enables while-drilling updating and correction of this confidence-based formation pressure interval profile, thereby more accurately obtaining formation pressure parameters. This provides more accurate formation pressure information for drilling risk assessment, reduces drilling risks associated with unclear formation pressure understanding, and effectively improves drilling efficiency.

[0142] The above is a detailed description of an embodiment of the present invention. However, the content described is only a preferred embodiment of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A method for estimating formation pore pressure while drilling based on Bayesian theory, characterized by: The following steps are included: S1: Collecting well logging data, logging while drilling data of the target well and data of adjacent wells in the same block as the target well; S2: Based on the well logging data and the logging while drilling data, predict the confidence interval profile of the target well and determine the prior probability of the formation pressure at any well depth. S2 includes the following steps: S21: Calculating the formation pressure of the entire well section of the target well based on the effective stress method formation pressure prediction model; S22: Substitute the formation pressure Pp of the entire well section calculated by the effective stress method into the DC index method formula, inversely calculate the Eaton index at any well depth along the vertical direction, and construct an analysis sample library of the Eaton index; S23: Using the sliding window method, the Eaton index at any well depth is converted from a single sample into a set of 2n+1 small samples. The normal information diffusion method is used for probability distribution fitting to establish the probability density function and cumulative probability density function of the Eaton index at any well depth. S24: Selecting a pure mudstone section as a normal compaction trend line construction interval, and defining a starting well section and an ending well section within the interval, wherein the starting well section contains n sample points, and the ending well section contains m sample points; S25: Selecting the nth sample point in the starting well section and the mth sample point in the ending well section respectively, constructing n×m normal compaction trend lines, and constructing an analysis sample library of slopes and intercepts of the normal compaction trend lines; S26: performing probability statistical analysis on the slope and intercept of the normal compaction trend line, selecting the normal information diffusion method for probability distribution fitting, and establishing a probability density function and a cumulative probability density function of the normal compaction trend line; S27: Generate a random number sample XN that conforms to the probability density function of the characteristic parameters of the Eaton index and the normal compaction trend line through Monte Carlo simulation, where N is the number of Monte Carlo simulations; S28: Substitute the random number samples generated in S27 into the DC index method calculation model to calculate the formation pressure random sample set P at any well depth position prnd , select normal distribution to fit the formation pressure to obtain the prior probability density function f(P p ) and the cumulative probability density function F(P p ); S29: Take the pressure values ​​of the formation pressure cumulative probability density J1 = 5% and J2 = 95% at any well depth, connect the points along the vertical well depth to form a line, and obtain the formation pressure curve with the cumulative probability of |J2-J1| = 90% for the entire well section of the target well, that is, the formation pressure interval profile with a confidence level of 90%; S3: Based on the adjacent well data, calculate the confidence interval profile of the adjacent well and determine the likelihood function of the formation pressure at any well depth; S4: Select the prior probability and likelihood function of the formation pressure at the same layer and substitute them into the Bayesian formula to obtain the updated posterior probability of the formation pressure.

2. The method for estimating formation pore pressure while drilling based on Bayesian theory according to claim 1, characterized in that: In S21, the effective stress method formation pressure prediction model is as follows: P P =S v -100.674e -2.5782μ Where: P P is the formation pressure, in MPa; S v is the overburden pressure, in MPa; μ is Poisson's ratio, dimensionless; Calculate the overburden pressure S by using the logging data v And the Poisson's ratio μ, the formula is as follows: Where: ρ is the average density of the overlying rock layer, in g / cm 3 ρ b is the formation density, in g / cm 3 ; H0 is the starting depth of the target well, in m; H is the ending depth of the target well, in m; Δt s is the shear wave time difference, unit is us / m; Δt P is the longitudinal wave time difference, unit is us / m.

3. The method for estimating formation pore pressure while drilling based on Bayesian theory according to claim 1 or 2, characterized in that: In the step S22, the calculation model of the Eaton index is as follows: Where: n is the Eaton index; d c is the measured d c Index; d cn The normal compaction trend line of the formation corresponds to d c Index; P h is the hydrostatic pressure, in MPa; d c The index is calculated using drilling data using the following formula: Where: T is drilling time, unit is min / m; N is rotary table speed, unit is r / min; W is drilling pressure, unit is KN; D b is the drill diameter, in mm; G h is the hydrostatic pressure gradient, in g / cm 3 ; ρ ECD is the circulating equivalent drilling fluid density, in g / cm 3 .

4. The method for estimating formation pore pressure while drilling based on Bayesian theory according to claim 1 or 2, characterized in that: In S23, the calculation model of the normal information diffusion method is as follows: Where: n is the sample size; h is the diffusion coefficient; x i is the observation sample; When the sample volume n is different, the diffusion coefficient h is calculated by the following formula: Where: b is the maximum value of the sample, b=max{x i }; a is the minimum value of the sample, a=min{x i }; σ is the sample standard deviation; ζ is the correlation coefficient, which is determined by the sample size.

5. The method for estimating formation pore pressure while drilling based on Bayesian theory according to claim 1 or 2, characterized in that: In said S28, the random sample set P of formation pressure at any well depth position prnd The calculation formula is as follows: Where: nrnd is the random number sample set of Eaton index; d cnrnd A sample set of random numbers for the normal compaction trend line.

6. The method for estimating formation pore pressure while drilling based on Bayesian theory according to claim 1 or 2, characterized in that: In the step S28, the prior probability density function f(P p ) and the cumulative probability density function F(P p ), the formula is as follows: Where: μ is the sample mean; σ is the sample standard deviation.

7. The method for estimating formation pore pressure while drilling based on Bayesian theory according to claim 1 or 2, characterized in that: Said S3 comprises the following steps, S31: According to the geological stratification of the block where the target well is located, screen and obtain information of neighboring wells in the same stratum as the target well; S32: According to S2, the likelihood function L(X) of the target layer formation pressure is obtained through the information of the adjacent wells. 1 |P).

8. The method for estimating formation pore pressure while drilling based on Bayesian theory according to claim 7, characterized in that: Said S4 comprises the following steps, S41: The prior probability density function f(P p ) and the likelihood function L(X 1 |P)Substitute into the Bayesian formula; S42: Obtain the posterior probability g(P|X 1 ) to update the formation pressure of the target well.

9. The method for estimating formation pore pressure while drilling based on Bayesian theory according to claim 8, characterized in that: The Bayesian formula is as follows: Where: p(X) is the marginal probability, that is, the normalization coefficient.

Images