CN113279747B - System and method for preparing drilling mud formula and performance parameters - Google Patents

Abstract

The invention provides a system and a method for allocating drilling mud formula and performance parameters, comprising a drilling stratum information module, an on-site drilling detection device, an expert system in the field of hydrologic well drilling, a machine learning prediction model and a control system; forming an automatic drilling mud formulation system; the on-site drilling detection device comprises a drilling parameter detection device, a well structure detection device and a mud performance parameter detection device; transmitting data to the expert system in the hydrologic well drilling field to judge the change condition of the stratum through a drilling stratum information module, a drilling parameter detection device and a well structure detection device, and providing a drilling mud formula of the corresponding stratum as an input parameter of a control system; the machine learning prediction model comprises a slurry performance parameter prediction model and a slurry formula additive addition prediction model. The invention can provide the corresponding slurry formula of the stratum according to the actual condition of the stratum on site, thereby greatly increasing the working benefit of drilling on site.

Description

A system and method for formulating drilling mud formula and performance parameters

technical field

The invention relates to the technical field of hydrological water well drilling engineering, in particular to a system and method for formulating drilling mud formula and performance parameters.

Background technique

Hydrological water well drilling is one of the important means of proving and exploiting groundwater resources. Especially in Qinghai Province, which is located in the plateau area, in recent years, with the acceleration of industrialization and urbanization in Qinghai Province, the per capita water consumption in Qinghai Province has been increasing year by year. Although the snowmelt in the mountains caused by global warming has alleviated the increasingly severe water demand in Qinghai Province to a certain extent; however, to solve the problem fundamentally, it is necessary to rely on hydrological drilling to seek to develop groundwater resources to solve the problem. A large number of groundwater resources are developed and utilized, and the groundwater level is declining year by year. The wells are getting deeper and deeper, and the formations encountered are becoming more and more complicated. This not only requires the development of a large number of new drilling technologies and methods, but also needs to apply these new technologies and methods to practical projects in a timely manner. The quality of mud preparation during construction determines the speed and quality of drilling work to a large extent. Engineers are also more and more aware that various complex situations in drilling engineering may be directly or indirectly related to all drilling operations. related to the drilling fluid used. Therefore, designing a reasonable mud formula is the key to successfully drilling hydrological water wells and reducing drilling costs.

However, as far as the current level of drilling mud technology is popularized in my country, most of the engineering construction process does not pay enough attention to mud design and use. Mud design and treatment agent selection are usually carried out by engineers and technicians with their own engineering experience. Yes, it was only after the drilling was blocked that I thought of contacting experts to adjust the mud on site. With the development of computer science, related technologies in the field of artificial intelligence are introduced into the design of mud formula. Various effective methods (expert experience, engineering experience, theoretical analysis, numerical calculation, experimental simulation, on-site monitoring) can be integrated into a knowledge-based intelligent system, and problems that are difficult to solve by one method can be transformed into another way to solve it. Moreover, in the fields of solid mineral exploration, engineering construction drilling, hydrological water well drilling and other fields, the acquisition of drilling mud expert experience and expert knowledge base is complicated and cumbersome, the development of design software is relatively backward, and the researched results have not been widely used. and promotion. Therefore, inheriting the existing research results, pioneering and innovating, and developing a drilling mud formula and performance parameter adjustment system and method that can be used on the drilling site, and provide effective assistance for the on-site technicians to correctly carry out the mud optimization design. A very meaningful thing.

Contents of the invention

In view of this, the present invention provides a system and method for formulating drilling mud formula and performance parameters, which solves the technical problem of providing high-performance drilling mud formula in real time during the drilling process of hydrological water wells, which cannot be satisfied in the prior art.

A drilling mud formulation and performance parameter allocation system, including: drilling formation information module, on-site drilling detection device, hydrological water well drilling field expert system, machine learning prediction model and control system;

The on-site drilling detection device includes: a drilling parameter detection device, a well body structure detection device and a mud performance parameter detection device;

The expert system in the field of hydrological water well drilling includes: an input unit, an expert reasoning module, an output unit and a whole-process informational human-machine interface;

The machine learning prediction model includes: a mud performance parameter prediction model and a mud formula additive dosage prediction model;

The whole-process informatized man-machine interface includes a predictive model database and an expert database;

The drilling formation information module, the drilling parameter detection device and the wellbore structure detection device are all connected to the input unit in the hydrological water well drilling expert system, and the output unit in the hydrological water well drilling expert system is connected to the A mud performance parameter prediction model in the control system and the machine learning prediction model;

The whole-process informatization man-machine interface is used to establish the prediction model database and the expert database, and complete the operations of adding, modifying, searching, and deleting the database in real time according to the actual drilling conditions of on-site drilling;

The output parameters of the mud performance parameter prediction model in the machine learning prediction model are compared with the mud performance parameter detection device, and the difference obtained is sent to the mud formula additive dosage prediction model, and the mud recipe additive dosage prediction model is connected to the control system.

Further, the drilling stratum information module is used to obtain stratum information in real time during the drilling process, including geological structure, rock properties and underground fluid conditions.

Further, the drilling parameter detection device is used for real-time detection of drilling parameters during the drilling process, including drilling pressure, rotational speed, pump pressure and drilling speed;

The well body structure detection device is used for real-time detection of well body structure parameters during the drilling process, including detection of the depth of the well section and the water content in the well.

Further, the mud performance parameter detection device is used to detect the mud circulation once in real time during the drilling process, and return the on-site mud performance parameters after sand removal, including density, pH value, apparent viscosity, plastic viscosity and dynamic shear force .

Further, the input unit is used to input a basic data source to the expert reasoning module; the basic data source includes formation information, drilling parameters detected and uploaded by the drilling parameter detection device, and wellbore structure detection device detected and uploaded. Uploaded shaft structure parameters.

Further, the expert reasoning module is used to read the basic data source input by the input unit, obtain formation conditions, drilling characteristics, and drilling well structure, and then obtain the formation category of the drilling formation, and judge the change of the drilling formation .

Further, the output unit is used to output the corresponding formation drilling mud formula; the corresponding formation drilling mud formula includes the composition and content of mud additives, and the formation drilling mud formula is sent to the control system to prepare the corresponding mud formula; At the same time, the corresponding formation drilling mud formula is used as the input parameter of the mud performance parameter prediction model.

Further, the mud performance parameter prediction model is used to receive the corresponding formation drilling mud formula delivered by the expert system in the field of hydrological well drilling, and predict the performance prediction parameters of the mud formula, including density, pH value, apparent viscosity, plastic viscosity and dynamic shear. The force is compared with the on-site mud performance parameters detected by the mud performance parameter detection device in real time.

Further, the mud formulation additive dosage prediction model is used to obtain the difference parameter between the mud formulation performance prediction parameters predicted by the mud performance parameter prediction model and the on-site mud performance parameters detected by the on-site mud performance parameter detection device in real time. The difference parameters include density, PH value, apparent viscosity, plastic viscosity and dynamic shear force, and predict the type and content of additives that need to be added to adjust the amount, and send them to the control system to correct the mud treated on site in time performance.

A method for formulating drilling mud formulas and performance parameters, applied to any one of the systems for formulating drilling mud formulas and performance parameters, comprising the following steps:

S1. Upload the drilling parameters, well section depth, downhole water content and formation information acquired by the input unit to the expert reasoning module in the hydrological water well drilling expert system;

S2. The expert reasoning module obtains the mud formula of the corresponding formation through rule-based reasoning, the mud formula includes the type and content of additives, and transmits the type and content information of additives to the control system to prepare on-site drilling mud;

S3. Input the mud formula obtained in step S2 into the machine learning prediction model, establish a mud performance parameter prediction model, and output corresponding mud performance parameters, including density, pH value, apparent viscosity, plastic viscosity, dynamic shear force ;

S4, compare the density, pH value, apparent viscosity, plastic viscosity, and dynamic shear force of the mud performance parameters obtained in step S3 with the field mud performance parameters detected by the mud performance parameter detection device, if at least one difference parameter is greater than the set threshold, then execute step S6, otherwise execute step S5;

S5. The expert reasoning module uses the formation conditions, drilling parameters and wellbore structure to infer and judge the change of the formation. If the formation changes, re-execute step S2 to obtain a new mud formula, otherwise directly use the on-site detection in step S4. Mud, continue to complete the drilling work;

S6. Input the difference parameters of mud performance parameters density, pH value, apparent viscosity, plastic viscosity and dynamic shear force obtained in step S4 into the mud formula additive dosage prediction model in the machine learning prediction model, and establish mud formula The additive dosage prediction model outputs the type and content of additives that need to be adjusted by dosage, and transmits the output information to the control system to timely correct the treated mud on site and continue to complete the drilling work.

The beneficial effects brought by the technical solution provided by the present invention are: (1) the present invention effectively solves the problem of large differences in the results of mud preparation due to differences in the technical level and working environment of different designers in the on-site drilling of hydrological water wells Or the problem that the ideal effect cannot be reached; (2) the present invention can provide the corresponding mud formula of the stratum according to the actual situation of the stratum on site, and further realize the mud addition treatment after the mud circulation treatment, thereby adjusting the drilling mud performance in time and improving The utilization rate of mud can greatly increase the work efficiency of on-site drilling and drilling, reduce unnecessary manpower and material resources, and provide scientific decision-making for the mud formula required for hydrological and water well drilling projects.

Description of drawings

Fig. 1 is a system architecture diagram of the present invention;

Fig. 2 is the architecture diagram of the expert system in the field of hydrological water well drilling of the embodiment of the present invention;

Fig. 3 is the frame diagram of the mud performance parameter prediction model of the embodiment of the present invention;

Fig. 4 is a framework diagram of a prediction model for adding additives in mud formula according to an embodiment of the present invention.

Detailed ways

In order to make the purpose, technical solution and advantages of the present invention clearer, the embodiments of the present invention will be further described below in conjunction with the accompanying drawings.

Please refer to Fig. 1, a drilling mud formulation and performance parameter allocation system provided by the present invention includes: drilling formation information module, on-site drilling detection device, hydrological water well drilling field expert system, machine learning prediction model and control system;

The on-site drilling detection device includes: a drilling parameter detection device, a well body structure detection device, and a mud performance parameter detection device;

The expert system in the field of hydrological water well drilling includes: an input unit, an expert reasoning module, an output unit and a whole-process informational human-machine interface;

The machine learning prediction model includes: a mud performance parameter prediction model and a mud formula additive dosage prediction model;

The whole-process informatized man-machine interface includes a predictive model database and an expert database;

The drilling formation information module, the drilling parameter detection device and the well body structure detection device in the on-site drilling detection device are all connected to the input unit in the hydrological water well drilling expert system, and the hydrological water well drilling field The output unit in the expert system is connected to the mud performance parameter prediction model in the control system and the machine learning prediction model;

The output parameters of the mud performance parameter prediction model in the machine learning prediction model are compared with the mud performance parameter detection device in the on-site drilling detection device, and the difference obtained is sent to the mud formula additive in the machine learning prediction model. A quantity prediction model, the mud formula additive dosage prediction model is connected to the control system.

The drilling formation information module is used to: obtain formation information in real time during the drilling process, including geological structure, rock properties, underground fluid conditions, etc.;

The drilling parameter detection device is used for: during the drilling process, real-time detection of drilling parameters, including drilling pressure, rotational speed, pump pressure, drilling speed, etc.;

The wellbore structure detection device is used for: during the drilling process, real-time detection of wellbore structure parameters, including detection of well section depth and downhole water content;

The mud performance parameter detection device is used for: during the drilling process, detect the mud circulation once in real time, and return the on-site mud performance parameters after desanding, including density, pH value, apparent viscosity, plastic viscosity, and dynamic shear force;

The input unit is used to: input a basic data source to the expert reasoning module; the basic data source includes the stratum information, the drilling data detected and uploaded by the drilling parameter detection device in the on-site drilling detection device. parameters, and the well structure parameters detected and uploaded by the well structure detection device in the on-site drilling detection device;

The expert reasoning module is used to: read the basic data source input by the input unit, obtain formation conditions, drilling characteristics, and drilling well structure, and then obtain the formation category of the drilling formation, and judge the change of the drilling formation Condition;

The output unit is used to: output the corresponding formation drilling mud formula; the corresponding formation drilling mud formula includes the composition and content of mud additives, which are sent to the control system to prepare the corresponding mud formula; at the same time, the corresponding formation drilling mud The formula is used as an input parameter of the mud performance parameter prediction model;

The mud performance parameter prediction model is used to: receive the corresponding formation drilling mud formula delivered by the expert system in the hydrological water well drilling field, and predict and obtain mud formula performance prediction parameters, including density, pH value, apparent viscosity, and plastic viscosity , dynamic shear force, compared with the on-site mud performance parameters detected by the mud performance parameter detection device in real time;

The mud formula additive dosage prediction model is used to: obtain the difference parameter between the mud formula performance prediction parameters predicted by the mud performance parameter prediction model and the on-site mud performance parameters detected by the mud performance parameter detection device in real time , the difference parameters include density, pH value, apparent viscosity, plastic viscosity, dynamic shear force, and predict the type and content of additives that need to be added to adjust the amount, and send them to the control system, and timely correct the on-site processed Mud performance;

Please refer to Figure 2. The input unit of the expert system based on hydrology and water well drilling is the formation conditions of the drilling formation, drilling parameters, and well structure. The expert reasoning module includes: reasoning engine, expert database and expert rule base, and the output unit is the mud formula. .

A method for formulating drilling mud formulas and performance parameters, applied to any one of the systems for formulating drilling mud formulas and performance parameters, comprising the following steps:

S1. The drilling parameter detection device and the well body structure detection device in the on-site drilling detection device are connected to the input unit in the hydrological water well drilling field expert system, and the input unit obtains the drilling parameters The drilling parameters detected by the detection device, the depth of the well section, the downhole water content and formation information detected by the well body structure detection device are uploaded to the expert reasoning module in the hydrological water well drilling expert system;

S2. The expert reasoning module in the hydrological water well drilling field expert system uses the acquired basic data including formation information (geological structure, rock properties, underground fluid conditions, etc.), drilling parameters (weight-on-bit, rotation speed, pump pressure, drilling speed, etc.), and the wellbore structure, the mud formula of the corresponding formation is obtained through rule-based reasoning. The control system is used to prepare the drilling mud, and the expert rules are expressed by production rules.

The logical expression of the expert rules and reasoning mechanism is as follows:

IF: formation structure, rock properties, underground fluid conditions, pressure on bit, speed, pump pressure, rate of penetration, and well structure

THEN: Drilling fluid type = formation mud and additive type and content = bentonite content + Na ₂ CO ₃ content + CaCl ₂ content + HEC content + LV-CMC content +...

Please refer to Fig. 3, the input parameter of mud performance parameter prediction model is the additive type and the content of the mud formula obtained in step S2, and described additive type and content include bentonite, soda ash, barite etc., output parameter is mud performance parameter, The output parameters are compared with the on-site mud performance parameters detected by the mud performance parameter detection device.

S3. Input the mud formula obtained in step S2 into the mud performance parameter prediction model in the machine learning prediction model, establish a mud performance parameter prediction model, and output corresponding mud performance parameters, including density, pH value, apparent viscosity , plastic viscosity, dynamic shear force;

S4, use the step S3 to obtain the density, pH value, apparent viscosity, plastic viscosity, dynamic shear force of the mud performance parameters, and the mud performance parameter detection device, and detect the density, pH value, apparent viscosity and plasticity of the mud performance parameters on site Viscosity, dynamic shear force and predicted parameters are compared, if there is at least one difference parameter greater than the set threshold, then execute step S6, otherwise execute step S5;

What the mud performance parameter detection device detects is the mud that has been recycled once in step S2 and returned to the desanding treatment;

S5. The expert reasoning module uses the formation conditions, drilling parameters and wellbore structure to infer and judge the change of the formation. If the formation changes, re-execute step S2 to obtain a new mud formula, otherwise directly use the on-site detection in step S4. Mud, continue to complete the drilling work;

Please refer to Fig. 4, the input parameter of the prediction model of the addition amount of the mud formula additive is the mud formula performance parameter density obtained in step S4, including the difference of density, pH value, apparent viscosity, plastic viscosity and dynamic shear force, and the output parameter is It is necessary to adjust the type and content of additives, including bentonite, soda ash, barite and other additives, and upload them to the control system;

S6. Input parameters to the prediction model of the addition of mud formula additives in the machine learning prediction model, and the input parameters are the mud performance parameters obtained in step S4. Density, pH value, apparent viscosity, plastic viscosity, dynamic shear Based on the force difference parameters, a prediction model for the addition of mud formula additives is established, and the type and content of additives that need to be adjusted are outputted, and the output information is transmitted to the control system to correct the treated mud in time and continue to complete the drilling work.

The input parameter of the mud formula additive amount prediction model is the difference parameter of the mud formula performance parameter, and the output parameter is the additive type and content of the mud formula additive amount; the two prediction models are the mud formula performance parameter model obtained by the forward push first , and then back-calculated to obtain the additive dosage model of the mud formula, and finally input the result of the reverse deduction of the additive dosage model of the mud formula into the performance parameter model of the mud formula, and compared the results of the two models to verify the accuracy of the model.

The whole-process informatization man-machine interface is used for: establishing the prediction model database and the expert database, which can facilitate on-site drilling personnel to complete the addition, modification, search, and deletion of the database in real time according to the actual drilling conditions of the on-site drilling Operation, so that the designed system is closer to the actual drilling work of hydrological water wells.

Embodiments of the present invention have been described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementations, and the above-mentioned specific implementations are only illustrative, rather than restrictive, and those of ordinary skill in the art will Under the enlightenment of the present invention, many forms can also be made without departing from the gist of the present invention and the protection scope of the claims, and these all belong to the protection of the present invention.

Claims (1)

1. A method of drilling mud formulation and performance parameter tuning, applied to a system of drilling mud formulation and performance parameter tuning, the system comprising:

drilling stratum information module, on-site drilling detection device, hydrologic well drilling field expert system, machine learning prediction model and control system;

the in situ drilling inspection apparatus includes: drilling parameter detection device, well structure detection device and mud performance parameter detection device;

the hydrologic well drilling field expert system includes: the system comprises an input unit, an expert reasoning module, an output unit and a whole process informationized human-computer interface;

the machine learning predictive model includes: a slurry performance parameter prediction model and a slurry formula additive addition prediction model;

the whole process informatization man-machine interface comprises a prediction model database and an expert database;

the drilling stratum information module, the drilling parameter detection device and the well structure detection device are all connected with an input unit in the hydrologic well drilling field expert system, and an output unit in the hydrologic well drilling field expert system is connected with the control system and the slurry performance parameter prediction model in the machine learning prediction model;

the whole process informatization man-machine interface is used for establishing the prediction model database and the expert database, and adding, modifying, searching and deleting the database are completed in real time according to the actual drilling condition of on-site drilling;

the output parameters of the slurry performance parameter prediction model in the machine learning prediction model are compared with the slurry performance parameter detection device, the obtained difference value is transmitted to a slurry formula additive addition prediction model, and the slurry formula additive addition prediction model is connected with the control system;

the input unit is used for inputting a basic data source to the expert reasoning module; the basic data source comprises stratum information, drilling parameters detected and uploaded by the drilling parameter detection device and well structure parameters detected and uploaded by the well structure detection device;

the expert reasoning module is used for reading the basic data source input by the input unit, obtaining stratum conditions, drilling characteristics and drilling well structures, further obtaining stratum types of drilling stratum, and judging the change condition of the drilling stratum;

the output unit is used for outputting a corresponding stratum drilling mud formula; the corresponding formation drilling mud formulation comprises the components and contents of mud additives, and the formation drilling mud formulation is conveyed into the control system to prepare the corresponding mud formulation; simultaneously taking the corresponding stratum drilling mud formula as an input parameter of a mud performance parameter prediction model;

the slurry performance parameter prediction model is used for receiving a corresponding stratum drilling slurry formula transmitted by an expert system in the field of hydrologic well drilling, predicting to obtain slurry formula performance prediction parameters including density, pH value, apparent viscosity, plastic viscosity and dynamic shear force, and comparing the slurry performance parameters with the field slurry performance parameters obtained by the slurry performance parameter detection device in real time;

the slurry formula additive addition prediction model is used for obtaining a difference parameter between a slurry formula performance prediction parameter predicted by the slurry performance parameter prediction model and a field slurry performance parameter detected by a field slurry performance parameter detection device in real time, wherein the difference parameter comprises density, pH value, apparent viscosity, plastic viscosity and dynamic tangential force, predicting the type and content of an additive to be added and adjusted, and transmitting the type and content to the control system to timely correct the field treated slurry performance;

the drilling stratum information module is used for acquiring stratum information including geological structures, rock properties and underground fluid conditions in real time in the drilling process;

the drilling parameter detection device is used for detecting drilling parameters in real time in the drilling process, wherein the drilling parameters comprise weight on bit, rotation speed, pump pressure and drilling speed;

the well structure detection device is used for detecting well structure parameters in real time in the drilling process, including detecting the depth of a well section and the water content in the well;

the mud performance parameter detection device is used for detecting mud circulation once in real time in the drilling process, and returning to the site mud performance parameters after desanding, including density, PH value, apparent viscosity, plastic viscosity and dynamic shear force, and is characterized by comprising the following steps:

s1, uploading drilling parameters, a well Duan Shendu, underground water content and stratum information acquired by an input unit to an expert reasoning module in an expert system in the hydrologic water well drilling field;

s2, an expert reasoning module obtains a mud formula of the corresponding stratum through a rule-based reasoning mode, wherein the mud formula comprises types and contents of additives, and information of the types and contents of the additives is transmitted to the control system to prepare on-site drilling mud;

s3, inputting the slurry formula obtained in the step S2 into the machine learning prediction model, establishing a slurry performance parameter prediction model, and outputting to obtain corresponding slurry performance parameters including density, pH value, apparent viscosity, plastic viscosity and dynamic shear force;

s4, comparing the density, the PH value, the apparent viscosity, the plastic viscosity and the dynamic shear force of the slurry performance parameters obtained in the step S3 with the on-site slurry performance parameters obtained by detection of the slurry performance parameter detection device, and executing the step S6 if at least one difference parameter is larger than a set threshold value, otherwise executing the step S5;

s5, an expert reasoning module uses stratum conditions, drilling parameters and well structures to reason and judge the change condition of the stratum, if the stratum changes, the step S2 is executed again, a new slurry formula is obtained, otherwise, the slurry detected in the field in the step S4 is directly used, and drilling work is continuously completed;

s6, inputting the differential value parameters of the slurry performance parameter density, the PH value, the apparent viscosity, the plastic viscosity and the dynamic shear force obtained in the step S4 into a slurry formula additive addition prediction model in the machine learning prediction model, establishing a slurry formula additive addition prediction model, outputting to obtain the types and the contents of additives required to be added and adjusted, transmitting output information to the control system, timely correcting the slurry treated on site, and continuously completing drilling work.

Images