Artificial Intelligence (AI) has found extensive usage in simplifying complex decision-making procedures in practically every competitive market field, and oil and gas upstream industry is no exception to it. AI involves the use of sophisticated networking tools and algorithms in solving multifaceted problems in a way that imitates human intellect, with the aim of enabling computers and machines to execute tasks that could earlier be carried out only through demanding human brainstorming. Unlike other simpler computational automations, AI enables the designed tools to "learn" through repeated operation, thereby continuously refining the computing capabilities as more data is fed into the system. Over the years, AI has led to significant designing and computation optimizations in the global Petroleum Exploration and Production (E&P) industry, and its applications have only continued to grow with the advent of modern drilling and production technologies. Tools such as Artificial Neural Networks (ANN), Generic Algorithms, Support Vector Machines and Fuzzy Logic have a historic connection with the E & P industry for more than 16 years now, with the first application dated in 1989 for development of an intelligent reservoir simulator interface, and for well-log interpretation and drill bit diagnosis through neural networks. Devices and softwares with basis from the above mentioned AI tools have been proposed to abridge the technology gaps hindering automated execution and monitoring of key reservoir simulation, drilling and completion procedures including seismic pattern recognition, reservoir characterisation and history matching, permeability and porosity prediction, PVT analysis, drill bits diagnosis, overtime well pressure-drop estimation, well production optimization, well performance projection, well / field portfolio management and quick, logical decision making in critical and expensive drilling operations. The paper reviews and analyzes this successful integration of AI techniques as the missing piece of the puzzle in many reservoir, drilling and production aspects. It provides an update on the level of AI involvement in service operations and the application trends in the industry. A summary of various research papers and reports associated with AI usage in the upstream industry as well as its limitations has been presented.
The process of drilling a borehole is very complex, involving surface and downhole drilling systems, which interact with the drilling fluid and the surrounding rocks. Modeling and simulating every aspect of the drilling process and drilling system is still considered too complex to be realized. However, many areas of modeling and simulation are currently undergoing very aggressive development. These areas include rig systems, downhole dynamics, rock-bit interaction, drilling/formation fluid, and the Earth model. High-fidelity models in these well-defined areas have demonstrated some success. Lately, drilling modelling and simulation has become one of the key factors for advancing drilling systems automation/control, intelligent managed pressure drilling and drilling optimization by understanding and/or predicting downhole dynamics. Modeling the magnitude and spatiotemporal distribution of uncertainty in the actual drilling process poses serious challenges in constructing high-fidelity models of the entire drilling system. However, the advancement of technology may dramatically improve the future of such an attempt to accurately model and simulate the whole drilling process. The topics presented in this paper include the current state of drilling process simulation software and simulators, the challenges of modelling drilling systems for automation and control, adaptive simulations for downhole drilling systems and the operator's perspective on drilling modelling. This paper examines the current state of drilling modelling and simulation and identifies its future goals.
A new tool (or better a new tool system) is showing up to face the "advanced wells" challenge in the next years. Having spent the first part of an extensive period of field testing, the system called "Rotary Closed-Loop System" (ROLS) looks much closer to being run in any extended reach, deep horizontal or complex multilateral well, also in an HP-HT environment. The ROLS was designed to automatically control the well geometry during directional drilling, even while rotating the drill string. The well path is adjusted by hydraulically powered expandable ribs which generate a radial steering contact force to the borehole wall. The amount and direction of the steering force is controlled by the integral downhole electronics, which are capable of steering the well to the desired direction. All parts of the system except a short steering sleeve work continuously in rotation. The ROLS may be operated with or without a downhole motor. Several sub-systems of ROLS have been field tested during 1994 and 1995, e.g., the hydraulic components, most electronic modules, and the bottom-to-surface communication. The first field trial of the complete unit was conducted in late 1995 in a special test well at Montrose, Scotland. Here, for the first time, the full scale of ROLS features was verified, such as the automated steering into any direction as required, the two-way communication link, and the programming of the downhole tool from the surface while drilling. On the basis of findings from this early field application, some technical changes were then made. The field test program was further continued in early 1996 to check ROLS drilling and directional performance with further broadened operating parameters. During the time at Montrose a total of 3,500 ft (1,067 m) have been drilled. The inclination was built from 50 to 710, and the total drilling time has been 354 hrs. All major functions of the system have been fully verified. One major application of the Rotary Closed Loop System will be to steer wells in extended horizontal sections when steerable motors are difficult to operate. Additionally, because of the elimination of sliding operation together with the precise course correction capability, conventional directional wells will also highly benefit in many cases. Introduction Drilling deep, highly deviated wells often requires the application of versatile systems for directional control.
Reduction of drilling costs in the oil and gas industry and the geothermal energy sector is the main driver for major investments in drilling optimization research. The best way to reduce drilling costs is to minimize the overall time needed for drilling a well. This can be accomplished by optimizing the non-productive time during an operation, and through increasing the rate of penetration (ROP) while actively drilling. ROP has already been modeled in the past using empirical correlations. However, nowadays, methods from data science can be applied to the large data sets obtained during drilling operations, both for real-time prediction of drilling performance and for analysis of historical data sets during the evaluation of previous drilling activities. In the current study, data from a geothermal well in the Hanover region in Lower Saxony (Germany) were used to train machine learning models using Random Forest™ regression and Gradient Boosting. Both techniques showed promising results for modeling ROP.
This papar was prepared for presentamn at the lS9d SPE European Petroleum Confefenca held m MlIan, Italy, 22-24 October 199STtus paper was selecled for pres6Waticmw an SPE Prqram Committee following review of information ccataimd in an abstract subm-,itbdby the author(s). Ccalen!s of the paper, as presented, have not ken rev!ewd by lhe Sc.oeiy CJPetroleum Engineers and we subjsct!0 correction~the author(s). The material, as presented, does not necessarily retlect any psition of the SocIely of Petroleum Engineers, its officers, or membels. Papers presemti al SPE mm3tlng8are subj8ct to publiMion revmw by Edtiorial Committees of tfm S&5efy of Petroleum En@wers, Permissionm copy Is reslncted to an abstracf d not more Ihan S00 wards, Illustrallons may not tm copied. The absfract should contain conspicuous acknowledgnmt of where and by tiom the pa$w was presented. Wriie Liixarian, SPE. PO. Box 8338S8, Rrhardscm, TX 75083-3s36, u.S A., fax Of -2 14-952-943&
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.