Artificial Intelligence (AI) is a new technological science that studies and develops theories, methods, technologies, and application systems for simulating, extending, and enhancing human intelligence. AI is also known as “smart machines” or “machine intelligence,” which refers to the intelligence exhibited by machines (mechanical devices) created by humans. Generally, artificial intelligence refers to the technology that enables machines to achieve goals that require human-like intelligence through artificially created computer program software.
What is Intelligent Drilling
In the industry, there are broad and narrow definitions of AI drilling. The broad definition of AI drilling not only refers to the automation of drilling, logging, and completion operations, but also encompasses all operations leading up to the discovery of oil reservoirs and enhancing production capacity. In contrast, the narrow definition of AI drilling refers specifically to the automation of drilling, logging, and completion operations themselves, commonly known as automated drilling.
From the above broad definition of AI drilling, it is evident that AI drilling encompasses various fields such as petroleum geology, geophysical exploration, logging, drilling, oil extraction, reservoir engineering, as well as mechanics, automation, and computer science. These fields were previously independent and non-continuous processes, but with AI, they require integration and collaboration to form an organic whole. Therefore, AI drilling aims to automatically adjust the design trajectory of the wellbore based on subsurface geological conditions and reservoir locations, and to automatically locate and drill into the optimal reservoir position to achieve maximum production capacity.
Intelligent Drilling Process
The first step is to provide geological reservoir characteristics, with geological and geophysical departments providing descriptions of geological conditions and the physical characteristics of the reservoir, forming a specific AI drilling expert system. The second step is real-time measurement and control during drilling, utilizing advanced drilling measurement and control technologies, employing the currently mature and developing 5W (MWD, LWD, SWD, PWD, and FEWD) as means, combined with directional drilling and downhole closed-loop control technologies, to collect various geological, stratigraphic, environmental information, and drilling status around the wellbore and in front of the drill bit in real-time. The third step is to accurately determine the target reservoir, where a powerful AI drilling computer system underground intelligently judges based on the latest available downhole data, combining it with previously known geological exploration data to discover the target oil and gas reservoir, accurately determining its location, size, shape, thickness, and orientation.
The fourth step is to optimize the best drilling process to achieve maximum production capacity as the objective function, optimizing various process parameters, and automatically drilling to find the best trajectory through the oil and gas layers. The fifth step is to provide direct actual data, giving firsthand data that closely reflects reality, such as reservoir description results and the actual wellbore trajectory and conditions of each section of the wellbore, and based on this, proposing preliminary recommendations for the best oil extraction methods.
However, each step in the above process requires the involvement and control of artificial intelligence, hence the term broad AI drilling. Practical experience has shown that the drilling process in petroleum engineering is a very complex and uncertain process. The information that can be collected from the drilling process and the geological environment is often imprecise, uncertain, and consists of approximate qualitative non-numerical random fuzzy information. Furthermore, it is necessary to conduct real-time rapid drilling control based on this random fuzzy information. Clearly, solving these challenges and achieving broad AI drilling will require further efforts in the future. Currently, the urgent task is to first develop automated drilling to realize narrow AI drilling.
Drilling Rig Operating System
The future AI drilling rig operating system located on the ground must achieve the following advanced functions: Continuous drilling, through the close cooperation of the rig’s lifting system and top drive device, allows for connecting a single drill pipe without stopping drilling or pumping, thus improving drilling efficiency. Continuous tripping, during the tripping process, instead of stopping to make up the drill string as conventional rigs do, AI robots will replace traditional rig workers to automatically complete the make-up of the drill string during tripping. Consequently, the speed of intelligent tripping is greatly increased; research results from Norway’s West Drilling Products Company indicate that the tripping speeds for different drill string diameters can reach 3600m/h, 2700m/h, and 1800m/h, significantly higher than the conventional rig’s tripping speeds of 600 to 900m/h.
Continuous casing and completion, during completion operations, whether casing or tubing, like tripping, does not require stopping to make connections, but instead is completed by AI robots, thus increasing operational speed; according to research results from Norway’s West Drilling Products Company, casing can reach speeds of 900m/h. Continuous circulation of drilling fluid, during tripping and drilling, the pumps do not stop, maintaining uninterrupted circulation of drilling fluid. This is because, as mentioned above, both drilling and tripping are continuous, allowing for continuous circulation of drilling fluid. Clearly, continuous circulation of drilling fluid is beneficial for pressure-controlled drilling, enhancing operational safety.
Double-wall drill string drilling, this method uses a double-wall drill string composed of an inner wall and outer wall, forming an annular space between the inner pipe and the outer wall. Under AI control, the double-wall drill string can achieve the following functions: Reverse circulation drilling, where drilling fluid is pumped down through the top drive and its rotary joint into the annular space between the inner wall and outer wall of the double-wall drill string, and then discharged from the drill bit’s nozzles, flowing upward into the annular space between the bottom drill tool assembly and the wellbore. When the fluid returns to the blowout preventer assembly, the double-wall drill string’s inner wall and the wellbore’s annular space are sealed off by a rotating control head, so the return drilling fluid along with rock cuttings can only enter the inner pipe of the double-wall drill string through the double float valve near the bottom and return to the surface. This implementation of reverse circulation drilling can eliminate the blockage of rock cuttings and protect oil and gas reservoirs. Power transmission and information transmission, the inner pipe of the double-wall drill string, after being insulated, can be used as a coaxial cable to supply power downhole; it can also transmit electrical information bidirectionally, with a data transmission rate of up to 64,000 bits per second, thus enabling large capacity data to be transmitted bidirectionally in real-time, which is very beneficial for promoting artificial intelligence.
Dual-gradient drilling, due to the higher density drilling fluid in the double-wall drill string and the lower density clean fluid (such as seawater) filling the annular space above the blowout preventer assembly, this difference in density creates two pressure gradients, thus enabling dual-gradient drilling, which plays a significant role in protecting oil and gas reservoirs and preventing wellbore collapse, especially important in deepwater drilling.
No riser drilling, in marine drilling technology, the riser, which surrounds the drill string, primarily serves to isolate seawater. In current conventional marine drilling practices, the lower end of the riser connects to the seabed wellhead, allowing the returning drilling fluid from the seabed to not mix with seawater above. However, when using the new type of double-wall drill string, the returning drilling fluid along with rock cuttings has already entered the inner pipe of the double-wall drill string through the double float valve near the bottom, thus isolating it from seawater, making the riser unnecessary, which significantly reduces drilling costs and improves economic efficiency.
Wide drilling fluid density window drilling, to prevent wellbore collapse, the upper limit of the pressure gradient formed by the drilling fluid density must not exceed the rock’s fracture strength; to prevent blowouts, the lower limit of the pressure gradient must not be lower than the pore pressure of the oil and gas reservoir. The difference between these upper and lower limits constitutes a so-called window, meaning that the drilling fluid’s pressure gradient must be maintained within this upper and lower limit range for safe drilling.
Reverse Circulation Drilling Technology
Currently, Norway’s Reelwell Company has launched a brand new double-wall drill string and its reverse circulation drilling technology, suitable for deepwater drilling in the North Sea, which is expected to be implemented soon.
The so-called “one-trip drilling” refers to completing all drilling tasks in one go without changing the drill bit to reach the drilling target. Therefore, the prerequisite for implementing “one-trip drilling” is the availability of super long-life drill bits. With the future development of new super long-life drill bits, two “one-trip drillings” can complete the task of drilling a well. The first “one-trip drilling” is for the surface well section, completing the tasks of setting surface casing and installing the wellhead device. The second “one-trip drilling” is to drill the remaining well section in one trip directly to the target oil and gas layer. Clearly, achieving “one-trip drilling” can significantly reduce drilling costs, especially in horizontal well drilling, its advantages become even more apparent.
The so-called “one-trip logging” refers to completing all logging requirements in one downhole trip. This technology can not only measure all required logging information during the drilling process but also collect downhole fluid and core samples, while also providing geological guidance and oil and gas reservoir descriptions during drilling. It is evident that achieving these tasks with “one-trip logging” will significantly reduce operational risks and improve the encounter rate and single-well output of oil and gas reservoirs. If the aforementioned “one-trip drilling” can be realized, then “one-trip logging” can be conducted simultaneously without occupying additional logging operation time, further simplifying the operational process and effectively reducing logging operation costs. However, the realization of “one-trip logging” is not an easy task, as it still needs to solve many challenges related to while-drilling logging technology, such as data transmission, storage, and so on.
Ultra-high temperature and pressure drilling, to cope with the high-temperature and high-pressure conditions downhole, requires the implementation of AI ultra-high temperature and pressure drilling. This technology’s realization mainly involves developing downhole instruments, tools, and materials that can withstand ultra-high temperatures and pressures, such as while-drilling measurement (MWD), while-drilling logging (LWD), and near-bit geological guidance instruments; directional tools, downhole batteries, drill bits, completion tools, and materials for drilling fluids, cementing, and downhole tubulars, etc. With continuous technological advancements, the issues of ultra-high temperature and pressure resistance for these tools are expected to be resolved in the near future. For instance, currently, foreign MWD and LWD, rotary steerable drilling systems, and screw drilling tools have achieved maximum temperature resistances of 200°C, 200°C, and 230°C, respectively, while the maximum temperature resistance for drilling fluids has reached 260°C, indicating that the realization of ultra-high temperature and pressure drilling is imminent. Currently, Norway’s West Drilling Products Company has developed an AI drilling rig operating system capable of achieving some of the above functions. However, the comprehensive realization of these functions still requires further development.
The downhole AI real-time guided safe drilling system can real-time measure and control the downhole drill bit, optimizing to achieve maximum production capacity while ensuring directional guidance for the drill bit, adopting the best process parameters for drilling, and automatically finding the optimal trajectory to accurately penetrate the oil and gas layers. Meanwhile, it must also conduct real-time status monitoring and accident diagnosis during drilling, promptly and safely handling accident exclusions to ensure safe drilling. To achieve these functions, this part consists of the following two expert systems.
AI Drilling Real-Time Control Expert System
The task of this expert system is to achieve automatic guidance of the drill bit, automatically select the best process parameters, and automatically find the optimal wellbore trajectory. Therefore, the core component of this system is a new type of intelligent drill bit. This drill bit consists of four parts: the sensor measurement part, the computer data processing and storage part, the power supply part, and the communication control part. The sensor measurement part is responsible for collecting real-time monitoring data obtained during drilling, such as direction, weight on bit, rotation speed, and other parameters.
The computer data processing and storage part stores geological conditions and reservoir characteristics information provided by the geological and geophysical departments; it can apply neural network methods to establish optimization models for drilling process parameters (such as weight on bit, rotation speed, etc.) with the objective function of maximizing oil and gas production capacity, and compare the computed optimal process parameters with the real-time data obtained during drilling, issuing instructions to adjust process parameters in real-time for automatic drilling optimization. The power supply part supplies power through high-speed transmission cables installed inside the new intelligent drill string (such as the aforementioned double-wall drill string), which also allows the drill bit to transmit information to surface personnel; the communication control part communicates directly with the surface via the transmission cable installed inside the drill string, enabling remote control.
Currently, foreign Apache Petroleum Technology Company has developed an intelligent drill bit that can autonomously consider and communicate directly with surface equipment, controlling drilling speed and direction. The new intelligent drill string developed by foreign oil service companies, which is equipped with high-speed data transmission cables, meets the requirements of intelligent drill bits.
AI Drilling Real-Time Monitoring and Fault Handling Expert System
The purpose of this system is to ensure safe drilling operations. Current technologies for diagnosing and handling drilling accidents and complex situations do not meet the actual needs on-site, often making it difficult to avoid unforeseen drilling anomalies. This newly created expert system employs neural network methods to establish models, creating a Bayesian network that encompasses various event types and their associated probabilities. This probabilistic model can utilize past and present data trends along with AI methods to predict drilling process accidents and equipment or sensor failures, with its network layout being used for real-time detection of various incidents related to well control and hydraulics, such as drill string leaks, pump failures, fluid losses, and packer failures. Additionally, the model can assess predicted trends, improving its prediction accuracy through self-learning and self-calibration, adjusting for distorted sensor data and model uncertainties, thereby ensuring prediction precision without worrying about false positives.
For example, in preventing blowout accidents, foreign companies have developed an AI blowout preventer hydraulic control device that utilizes advanced PLC and touch screens, along with highly reliable PROFIBUS bus systems, to predict in real-time and issue timely alarms. Moreover, to detect potential drill string leaks and pump failures and initiate alarms, foreign companies have developed a new method that uses real-time monitoring data of drilling fluid flow speed trends and simulated pump pressure and flow correlations to jointly describe equipment status for evaluation; during extreme conditions, real-time alarms can be issued. The first phase of the alarm can determine the risk of leaks or pump failures and the potential time of occurrence; subsequently, leaks and pump failures are treated as a whole rather than individually identified, allowing for precise identification of the fault type represented by the alarm in the second phase.
The aforementioned downhole AI real-time guided safe drilling system has been used on an offshore drilling platform in North America. During a drilling operation on the platform, the system issued six alarms before a drill string leak occurred. As a result, the operators were able to take timely remedial action, eliminating catastrophic operational failures and preventing larger-scale economic losses. Furthermore, due to the system’s early event detection and swift problem resolution, non-productive time was significantly reduced.
Drilling Site AI Control Console Using Advanced Integrated AI Technology
This technology integrates comprehensive control over various aspects of pattern recognition, parameter optimization, system optimization, and effect prediction in oil and gas well drilling engineering. Its remote control method is through electrical signals, connected via multi-core cables, with a fast response speed, capable of reflecting the position of the valve handle in real-time, with speed and precision superior to traditional pneumatic control methods. The main function of the drilling site AI control console is to “liberate the driller”; that is, to replace the driller and complete all control tasks, allowing the driller to be freed from complex and tense operations, no longer needing to be on-site for extended periods, only required for special operations at the site.
A drilling site AI control console includes components such as a rack, explosion-proof plug, explosion-proof touch computer, explosion-proof buzzer alarm, explosion-proof buttons, and explosion-proof rotary switches. The upper part of the rack is equipped with an explosion-proof box, which contains an explosion-proof touch industrial computer that runs software for real-time guided control drilling and real-time monitoring and fault handling systems, such as shut-in system software, etc. The touch screen features various switch interfaces for individual blowout preventers in the blowout preventer assembly, valve switch interfaces for the choke manifold, parameter control interfaces, one-click shut-in interfaces, BOP and choke manifold custom combination interfaces, and data curve and control record interfaces. The left side of the touch screen has an explosion-proof main control button to start the computer program, while the right side has an explosion-proof rotary switch; at the bottom of the explosion-proof box, there is an explosion-proof plug and an explosion-proof buzzer, with the explosion-proof plug connected to the input and output terminals of the explosion-proof industrial computer, and the other end connected via multi-core cables to the existing blowout preventer control system PLC. The switch interface for individual blowout preventers visually displays the arrangement of blowout preventers and allows for custom combinations; the valve switch interface for the choke manifold visually displays the combination of the choke manifold and allows for custom combinations; both interfaces can be customized based on actual conditions to meet the needs of different wellheads and control systems. The one-click shut-in interface can be customized based on actual wellhead conditions, and when the one-click shut-in program runs, it can display shut-in steps and guide the next steps through indications. The explosion-proof industrial computer connects to the blowout preventer control system PLC via multi-core cables, used for transmitting pressure data from pressure sensors on the blowout preventer and choke manifold, alarm signal transmission, and control valve switching, with real-time data collection and record-keeping.
Taking the control shut-in as an example, the implementation process is as follows. When the explosion-proof industrial computer loads the shut-in system software, simply clicking on the control program on the explosion-proof touch industrial computer will convert the touch screen signal into a communication signal and transmit it via cable to the blowout preventer control system PLC, which then analyzes the signal and converts it into an electrical signal to control the solenoid valves on the choke manifold and other functions, while the touch screen displays the electrical signal returned from the PLC, showing the valve’s opening and closing position and pressure data in real-time. If there is an abnormal pressure at the wellhead requiring an emergency shut-in, simply clicking the start button on the one-click shut-in interface on the touch screen will automatically execute the shut-in procedure, guiding manual shut-in steps through prompts. It can also record all shut-in steps, pressure data, etc., for future reference. Clearly, such automation of the shut-in procedure greatly shortens shut-in time, reduces human error, and improves economic efficiency.
Remote AI Control Center. The distribution area of oil and gas wells in oil fields, whether onshore or offshore, is generally very dispersed and vast. To ensure the normal operation of numerous oil and gas well drilling tasks and achieve artificial intelligence, it is necessary to transmit images, drilling completion processes, logging data, and equipment operation and maintenance statuses in real-time to the oil field management department, allowing them to accurately grasp reliable first-hand intelligence and make timely correct judgments and command dispatches, feeding back to each oil and gas well in real-time, thus requiring the establishment of a remote AI control center. Typically, the structure of the remote control center for AI drilling consists of the following:
Core devices, mainly consisting of data servers and dedicated lines. The server installs operating system software, database software, monitoring system software, etc. These software systems have multiple functions, including active inquiries, data display, data storage, data querying, alarm display, generating current, voltage, and indicator curve graphs, etc.
Communication platform, which can typically use wireless LAN, data radio, GPRS, and other communication methods. Using digital network wireless transmission technology for data communication has advantages such as quick installation and opening, convenient maintenance and migration, low cost, and easy centralized management, especially when using GPRS, which not only reduces the number of on-site personnel but also enables remote joint control of various oil and gas well drilling rigs. Earth satellite communication (GPRS) includes: data analysis server clusters and deep learning server clusters. The former is used to analyze real-time data sent from controlled systems during the drilling control process of various oil and gas wells, obtaining different types of real-time analysis data and sending them to the deep learning server cluster; the latter combines real-time analysis data with historical analysis data for machine learning to obtain optimization data and sends it to the identified data analysis server cluster; the identified data analysis server cluster can then generate optimization instructions for the drilling control process of the controlled oil and gas wells based on the identified optimization data and send them to the controlled system.
Monitoring equipment, the monitoring equipment of the remote AI control center mainly includes: remote measurement and control terminals for oil and gas wells, signal/data acquisition devices for well working parameters, on-site power modules, on-site data display, operation terminals, and sealed explosion-proof electrical boxes, etc. Among them, the remote measurement and control terminal consists of a remote control terminal, power supply, front-end sensors, and protection boxes, while the remote control terminal comprises data acquisition I/O, power modules, communication modules/interfaces, antennas, etc.; the protection box is a casing to protect the measurement and control system, ensuring it is rainproof, sunproof, and dustproof. The on-site data display and operation terminals in the control center are equipped with LCD screens, with operational buttons on the panel to enable configuration and data display of on-site equipment.
Measurement equipment, including tension transmitters, angular displacement transmitters, current and voltage transformers, etc.
Currently, in China, Phoenix Technology Yield Fund Company has developed a remote drilling system utilizing satellite transmission; Weikong Technology Company has developed a remote monitoring system for oil wells centered on remote measurement and control terminals (RTU), creating favorable conditions for the establishment of remote control centers for AI drilling.
Globally, foreign oil service and technology companies have gradually launched some products related to AI drilling, and it is expected that by 2025, AI drilling will enter its initial stage, ushering in a new era of AI drilling, and the vision of “unmanned drilling (no personnel required at the drilling site)” will also be realized in the not-too-distant future. Therefore, AI drilling represents a comprehensive and profound revolution in current drilling technology, which will have far-reaching impacts on the drilling industry and its practitioners, significantly enhancing drilling efficiency, quality, safety, reliability, and economic and social benefits.
Thus, it is evident that China must catch up and vigorously develop AI drilling research and development with a spirit of urgency, transitioning from “catching up” to “leading” within a short period, fully entering the new era of AI drilling.
Information sourced from Petroleum and Equipment
