CN112796746B - A drilling method for petroleum geological exploration - Google Patents

Abstract

The invention discloses a drilling method for petroleum geological exploration, which comprises the following steps: s1, acquiring petroleum geological exploration historical detection data based on geological cloud, preprocessing the acquired historical detection data, and constructing a historical detection data set; s2, constructing a geological type recognition model based on the convolutional neural network, and training the geological type recognition model by adopting a historical detection data set; s3, drilling after connecting a plurality of optical fiber distributed sensors on the drilling device, acquiring corresponding parameter data of a well exploration and a drill bit in real time through the optical fiber distributed sensors, and acquiring formation parameters based on the real-time acquired well exploration parameter data; and S4, inputting the stratum parameters acquired in real time into the trained geological type recognition model to obtain the geological type distribution of the stratum, and adjusting the angle of the drill bit in real time based on the geological type distribution of the stratum to finish drilling. The invention can effectively improve the drilling efficiency, shorten the period of petroleum geological exploration and reduce the exploration cost.

Description

A drilling method for petroleum geological exploration

technical field

The invention relates to the technical field of petroleum geological exploration, in particular to a drilling method for petroleum geological exploration.

Background technique

Oil is one of the most important energy sources in the world. With the development of global society and economy and the improvement of scientific and technological level, the demand for oil in the entire international community is increasing. my country's oil extraction volume cannot meet the needs of domestic social development, which brings great pressure to the innovation and development of my country's oil exploration technology, and the subsequent development of oil resources also doubles the pressure. In order to ensure the sustainable and healthy development of my country's national economy and social productive forces, and to ensure the continuous exploitation of my country's petroleum resources, it is necessary to strengthen and innovate the exploration technology of petroleum geology, and comprehensively improve the quality and efficiency of my country's petroleum resources exploitation. Therefore, the innovation of oil exploration technology has become an inevitable development trend of the entire oil industry.

However, the traditional petroleum geological exploration technology has great defects in terms of maximizing oil exploitation and investment funds, and with the development of society and economy, the quality of petroleum exploration and development is getting smaller and smaller. The disadvantages of exploration technology are becoming more and more obvious, and the continuous innovation of petroleum geological exploration technology has become a necessity for the development of the times. Especially in terms of drilling, because the oil exploration and development process is composed of many stages of different nature and different tasks, and the purpose and tasks of drilling are different in different stages, traditional petroleum geological exploration technology is used in the whole oil exploration and development. During the process, the wells that need to be drilled include: reference wells, profile wells, parametric wells, structural wells, exploration wells, data wells, production wells, water injection wells, inspection wells, observation wells, and adjustment wells, which not only prolongs the exploration cycle, but also greatly increases the number of wells. Greatly increased the cost of petroleum geological exploration.

With the development of information technologies such as big data, cloud computing, and the Internet of Things, it is particularly necessary to integrate artificial intelligence with address exploration to provide a drilling method for petroleum geological exploration.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a drilling method for petroleum geological exploration, so as to solve the technical problems existing in the prior art, effectively improve drilling efficiency, shorten the period of petroleum geological exploration, and reduce exploration costs.

To achieve the above object, the present invention provides the following scheme: the present invention provides a drilling method for petroleum geological exploration, comprising:

S1. Obtain historical exploration data of petroleum geological exploration based on the geological cloud, and preprocess the acquired historical exploration data to construct a historical exploration data set;

S2, constructing a geological type identification model based on a convolutional neural network, and using historical detection data sets to train the geological type identification model;

S3. Drilling is performed after connecting several optical fiber distributed sensors on the drilling device, and the corresponding parameter data of the exploratory well and the drill bit are collected in real time through the optical fiber distributed sensors, and the formation parameters are obtained based on the real-time collected exploratory well parameter data;

S4. Input the real-time acquired stratum parameters into the trained geological type identification model to obtain the geological type distribution of the stratum, and adjust the angle of the drill bit in real time based on the geological type distribution of the stratum to complete the drilling.

Preferably, in the step S1, the acquired historical exploration data of petroleum geological exploration includes: detection parameter data and geological types corresponding to the detection parameter data.

Preferably, the detection parameter data includes: resistivity, permeability, rock density, fluid density, and water content of the formation.

Preferably, in the step S1, the data preprocessing method includes: outlier elimination and data normalization.

Preferably, in the step S2, in the training process of the geological type identification model, a stochastic gradient descent algorithm is used to update the parameters of the convolutional neural network.

Preferably, in the step S3, the optical fiber distributed sensor includes: a resistivity sensor, a stress sensor, an acoustic wave sensor, a liquid density sensor, a dielectric constant sensor, a temperature sensor, an angle sensor, and a laser sensor; wherein,

The resistivity sensor is used to collect the resistivity of the formation; through the collected resistivity, the water layer and the oil layer can be accurately judged;

The stress sensor is used to collect the stress of the formation;

The acoustic wave sensor is used for collecting acoustic wave data in the exploration well;

The liquid density sensor is used to collect the liquid density in the formation;

The dielectric constant sensor is used to collect the water content of the formation;

The temperature sensor is used to collect the borehole wall temperature of the exploratory well;

The angle sensor is used to collect the angle between the axis of the drill bit and the direction perpendicular to the ground;

The laser sensor is used to collect the distance between the bottom of the well and the ground, that is, the depth of the exploratory well.

Preferably, in the step S3, the method for fixing the optical fiber distributed sensor includes:

First, splicing the optical fiber distributed sensor with the optical fiber, and reeling the optical fiber onto the cable car of the drilling device;

Next, a stretching block is suspended from one end of the optical fiber extending into the exploratory well.

Preferably, in the step S4, the adjustment method of the drill bit angle specifically includes:

The formation parameters obtained from the real-time exploratory well data are input into the trained geological type identification model, and the geological types at different times are obtained. Obtain the best drilling direction, and adjust the angle of the drill bit in real time according to the angle between the axis of the drill bit and the vertical ground direction.

The present invention discloses the following technical effects:

(1) The present invention builds a geological type identification model based on a convolutional neural network, and obtains historical detection data through the geological cloud to train the geological type identification model, thereby realizing the organic integration of big data and intelligent petroleum geological exploration. The type identification model can realize accurate prediction of geological exploration types, thereby improving drilling efficiency and reducing exploration cycle;

(2) The present invention collects multiple well logging data in real time while drilling through optical fiber distributed sensors, and realizes fast and effective data transmission through optical fiber, without considering the problem of downhole data storage in the process of collecting a large amount of data; The type identification model can quickly and accurately predict the data collected by optical fiber distributed sensors, effectively reducing the number of wells, thus greatly shortening the period of petroleum geological exploration, and reducing exploration costs;

(3) The present invention adjusts the angle of the drill bit in real time through the distribution of geological types of the stratum, which can effectively ensure the quality of oil production and improve the drilling efficiency.

Description of drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings required in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some of the present invention. In the embodiments, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without any creative effort.

Fig. 1 is the flow chart of the drilling method used for petroleum geological exploration according to the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

In order to make the above objects, features and advantages of the present invention more clearly understood, the present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments.

1, the present embodiment provides a drilling method for petroleum geological exploration, comprising the following steps:

S1. Obtain historical exploration data of petroleum geological exploration based on the geological cloud, and preprocess the acquired historical exploration data to construct a historical exploration data set;

The geological cloud is a comprehensive geological information service system developed under the auspices of the China Geological Survey. It adopts a classic 4-layer cloud architecture and integrates four subsystems of geological survey, business management, data sharing and public services. It is oriented to geological survey technology. Personnel provide an intelligent geological survey work platform in the cloud environment, and innovate a new mode of geological survey work; for geological survey managers, provide one-stop comprehensive business management in the cloud environment and auxiliary decision support under the support of big data to realize geological survey projects, personnel, One-stop service for finance, equipment, etc.; provide various professional data sharing services such as basic geology, mineral geology, hydraulic environment geology, marine geology, etc. for all kinds of geological survey professionals; provide various types of geological information product services for the public .

The acquired historical exploration data of petroleum geological exploration includes: exploration parameter data and geological types corresponding to the exploration parameter data; and the exploration parameter data includes: formation resistivity, permeability, rock density, fluid density, and water content.

The data preprocessing method includes:

Hierarchical agglomerative clustering method based on average distance is used to remove outliers in historical detection data, and the remaining data after removal is normalized to reduce the interference caused by different dimensions of dependent variables.

S2, constructing a geological type identification model based on a convolutional neural network, and using historical detection data sets to train the geological type identification model;

The convolutional neural network includes an input layer, a convolutional layer, an activation layer, an output layer, and an auxiliary layer; the input layer is used to input the detection parameter data in the historical data set;

The convolution layer is used for feature extraction on the input detection parameter data;

The auxiliary layer is used to reduce the transition fitting of data, which helps to improve the ability of the network to generate training data and reduce training time; the auxiliary layer includes a truncation layer and a batch normalization layer; the stage layer is used to truncate some neurons The batch normalization layer is used to standardize the batch input data, reduce the dependence of the neural network learning on the parameter initialization method, make the input data distribution stable, and accelerate the convergence rate of the neural network. , thereby speeding up the training of the neural network.

The output layer is used to output the geological type; the stochastic gradient descent algorithm is used to update the value of the parameters of the convolutional neural network, and the value of the loss function is reduced, so that the predicted geological type and the actual geological type gradually converge.

S3. Drilling is performed after connecting several optical fiber distributed sensors on the drilling device, and the corresponding parameter data of the exploratory well and the drill bit are collected in real time through the optical fiber distributed sensors, and the formation parameters are obtained based on the real-time collected exploratory well parameter data;

The optical fiber distributed sensor includes: resistivity sensor, stress sensor, acoustic wave sensor, liquid density sensor, dielectric constant sensor, temperature sensor, angle sensor, laser sensor;

Wherein, the resistivity sensor is used to collect the resistivity of the formation; through the collected resistivity, the water layer and the oil layer can be accurately judged;

The stress sensor is used to collect the stress of the formation; based on the corresponding relationship between the formation stress and the permeability, the permeability of the formation can be effectively calculated through the collected formation stress;

The acoustic wave sensor is used to collect the acoustic wave data in the exploration well; based on the relationship between the acoustic wave propagation speed and the rock density, the acquired acoustic wave data can effectively calculate the rock density of the formation;

The liquid density sensor is used to collect the liquid density in the formation;

The dielectric constant sensor is used to collect the water content of the formation;

The temperature sensor is used to collect the borehole wall temperature of the exploratory well; during the drilling process, the temperature near the bottom of the well is very high, and the thermal stress generated by the change of the borehole wall temperature during the drilling process will cause the formation stress near the borehole wall to change. Therefore, Through the real-time monitoring of the borehole wall temperature, when the borehole wall temperature exceeds the preset threshold, the exploratory well is cooled by the circulation of the drilling fluid, which can effectively ensure the accuracy and validity of the collected exploratory well parameters.

The angle sensor is used to collect the angle between the axis of the drill bit and the direction perpendicular to the ground, and through the angle, the direction of the drill bit can be adjusted;

The laser sensor is used to collect the distance between the bottom of the well and the ground, that is, the depth of the exploratory well; by collecting the distance between the bottom of the well and the ground, not only can the depth of the exploratory well be acquired to provide data support for subsequent oil production, but also the distributed distribution can be adjusted in real time according to the depth of the well. Fiber optic sensor to ensure that the distributed fiber optic sensor can be consistent with the position of the drill bit.

The fixing method of the optical fiber distributed sensor includes:

First, the optical fiber distributed sensor is spliced with the optical fiber, and the optical fiber is coiled onto the cable car of the drilling device; wherein, the optical fiber is wrapped in the optical fiber cable to prevent the harsh environment in the exploration well from causing the optical fiber to be damaged. damage;

Secondly, a stretch block is suspended from one end of the optical fiber extending into the exploration well, so that the optical fiber distributed sensor moves down together with the drill bit, so as to ensure that the distributed optical fiber sensor can keep the same position as the drill bit; at the same time, the bending of the optical fiber can be effectively avoided Or knotted to ensure the effective transmission of data collected by optical fiber distributed sensors.

Through optical fiber transmission, the simultaneous transmission of data collected by multiple optical fiber distributed sensors can be realized without considering the problem of downhole data storage.

S4. Input the real-time acquired stratum parameters into the trained geological type identification model to obtain the geological type distribution of the stratum, and adjust the angle of the drill bit in real time based on the geological type distribution of the stratum to complete the drilling.

Specifically: input the stratum parameters obtained from the exploratory well data collected in real time into the trained geological type identification model to obtain the geological types at different times; According to the angle of the axis of the collected drill bit and the vertical ground direction, the angle of the drill bit can be adjusted in real time, so as to effectively ensure the oil production quality and improve the drilling efficiency.

The present invention has the following technical effects:

(1) The present invention builds a geological type identification model based on a convolutional neural network, and obtains historical detection data through the geological cloud to train the geological type identification model, thereby realizing the organic integration of big data and intelligent petroleum geological exploration. The type identification model can realize accurate prediction of geological exploration types, thereby improving drilling efficiency and reducing exploration cycle;

(2) The present invention collects multiple well logging data in real time while drilling through optical fiber distributed sensors, and realizes fast and effective data transmission through optical fiber, without considering the problem of downhole data storage in the process of collecting a large amount of data; The type identification model can quickly and accurately predict the data collected by optical fiber distributed sensors, effectively reducing the number of wells, thus greatly shortening the period of petroleum geological exploration, and reducing exploration costs;

(3) The present invention adjusts the angle of the drill bit in real time through the distribution of geological types of the stratum, which can effectively ensure the quality of oil production and improve the drilling efficiency.

The above-mentioned embodiments are only to describe the preferred modes of the present invention, but not to limit the scope of the present invention. Without departing from the design spirit of the present invention, those of ordinary skill in the art can make various modifications to the technical solutions of the present invention. Variations and improvements should fall within the protection scope determined by the claims of the present invention.

Claims (4)

1. a drilling method for petroleum geological exploration, is characterized in that, comprises the following steps: S1. Obtain historical exploration data of petroleum geological exploration based on the geological cloud, and preprocess the acquired historical exploration data to construct a historical exploration data set; The preprocessing method includes: outlier elimination and data normalization; S2, constructing a geological type identification model based on a convolutional neural network, and using historical detection data sets to train the geological type identification model; The convolutional neural network includes an input layer, a convolutional layer, an activation layer, an output layer, and an auxiliary layer; the convolutional layer is used to perform feature extraction on the input detection parameter data; the auxiliary layer is used to reduce the transition of data fitting; the auxiliary layer includes a truncation layer and a batch normalization layer; the truncation layer is used to truncate the connections between some neurons; the batch normalization layer is used to perform a normalization operation on the batch input data to reduce neural network learning Dependence on parameter initialization method; the output layer is used to output geological type; During the training process of the geological type identification model, a stochastic gradient descent algorithm is used to update the parameters of the convolutional neural network; S3. Drilling is performed after connecting several optical fiber distributed sensors on the drilling device, and the corresponding parameter data of the exploratory well and the drill bit are collected in real time through the optical fiber distributed sensors, and the formation parameters are obtained based on the real-time collected exploratory well parameter data; The optical fiber distributed sensor includes: resistivity sensor, stress sensor, acoustic wave sensor, liquid density sensor, dielectric constant sensor, temperature sensor, angle sensor, laser sensor; wherein, The resistivity sensor is used to collect the resistivity of the formation; through the collected resistivity, the water layer and the oil layer can be accurately judged; The stress sensor is used to collect the stress of the formation; The acoustic wave sensor is used for collecting acoustic wave data in the exploration well; The liquid density sensor is used to collect the liquid density in the formation; The dielectric constant sensor is used to collect the water content of the formation; The temperature sensor is used to collect the borehole wall temperature of the exploratory well; The angle sensor is used to collect the angle between the axis of the drill bit and the direction perpendicular to the ground; The laser sensor is used to collect the distance between the bottom of the well and the ground, that is, the depth of the exploratory well; The fixing method of the optical fiber distributed sensor includes: First, splicing the optical fiber distributed sensor with the optical fiber, and reeling the optical fiber onto the cable car of the drilling device; secondly, extending the end of the optical fiber into the exploration well and suspending the stretching block; S4. Input the real-time acquired formation parameters into the trained geological type identification model to obtain the geological type distribution of the formation, and adjust the angle of the drill bit in real time based on the geological type distribution of the formation to complete the drilling. 2. The drilling method for petroleum geological exploration according to claim 1, characterized in that, in the step S1, the acquired historical exploration data of petroleum geological exploration comprises: detection parameter data and the detection parameter data corresponding to the detection parameter data. the corresponding geological type. 3 . The drilling method for petroleum geological exploration according to claim 2 , wherein the detection parameter data comprises: resistivity, permeability, rock density, fluid density and water content of the formation. 4 . 4. The drilling method for petroleum geological exploration according to claim 1, wherein in the step S4, the adjustment method of the drill bit angle specifically comprises: The formation parameters obtained from the real-time exploratory well data are input into the trained geological type identification model to obtain the geological types at different times, and the geological type distribution of the formation is obtained based on the depth of the exploration wells corresponding to different times. Based on the geological type distribution of the formation, Obtain the best drilling direction, and adjust the angle of the drill bit in real time according to the angle between the axis of the drill bit and the vertical ground direction.

Images