Artificial intelligence is used to capitalize on commonly run well logs to build and train an artificial neural network model to predict the sonic log curve. The dataset used to develop the model is made up of a total of 50 well logs for an oilfield that is located in the eastern part of US. It includes actual recorded sonic logs, which allows a comparison between model driven results and actual sonic curves to evaluate accuracy. Commonly run logs are used to train prediction models. Log selection targets meaningful petrophysical correlations to sonic properties. Well logs were preprocessed for data cleanup and scaling. After that, the data was broken down to geologically continuous chunks, and splitted to training, validation, and testing subsets. Artificial neuronal network models that use different combinations of neural network layers, hyperparameters, optimizers, and architectures to train the model are investigated to decide on the best approach to construct the predictive model. Automated hyperparameters tuning is employed to further enhance the model accuracy. After evaluation, a total of four models are selected to train. The resulted sonic logs produced by the models are assessed for accuracy, and are visually plotted and compared to actual logs. The models are ranked based on their accuracy and ability to detect geological features. The models show variations in prediction capabilities. They show that models that are built based on recurrent neural network performed the best. In addition, introducing convolutional neural layers to the models further enhanced accuracy. It is observed that the models are doing a better job of predicting when there is a clear change in values, and are less effective when actual values rate of change is low. Overall, the model that is built employing Gated Recurrent Unit combined with convolutional layers performed the best. In comparison, models that are purely built on dense neural network structure didn't yield good results. Also, it is proven that automated hyperparameter functions are effective in improving models’ accuracy by tuning of parameters through several iterations. The paper exhibits the effectiveness of utilizing evolving AI techniques to capitalize on expanding existing formation evaluation methods to extract more data from commonly run well logs. In addition, developing a model to predict sonic response negates the need to run actual sonic logs, which saves cost and operational time. Also, AI derived data from previously acquired logs in the wells of interest can help in calibration of models in applications such as geomechanics and wellbore stability studies. This approach can be potentially used for real-time predications without running sonic tools in wells.
Monitoring and surveillance (M&S) is one of the key requisites for assessing the effectiveness and success of any Improved Oil Recovery (IOR) or Enhanced Oil Recovery (EOR) project. These projects can include waterflooding, gas flooding, chemical injection, or any other types. It will help understand, track, monitor and predict the injectant plume migration, flow paths, and breakthrough times. The M&S helps in quantifying the performance of the IOR/EOR project objectives. It provides a good understanding of the remaining oil saturation (ROS) and its distribution in the reservoir during and after the flood. A comprehensive and advanced monitoring and surveillance (M&S) program has to be developed for any given IOR/EOR project. The best practices of any such M&S program should include conventional, advanced and emerging novel technologies for wellbore and inter-well measurements. These include advanced time-lapse pulsed neutron, resistivity, diffusion logs, and bore-hole gravity measurements, cross-well geophysical measurements, water and gas tracers, geochemical, compositional and soil gas analyses, and 4D seismic and surface gravity measurements. The data obtained from the M&S program provide a better understanding of the reservoir dynamics and can be used to refine the reservoir simulation model and fine tune its parameters. This presentation reviews some proven best practices and draw examples from on-going projects and related novel technologies being deployed. We will then look at the new horizon for advanced M&S technologies.
Pump and pull cementing method is introduced to overcome soft cement plugs due to the swabbing effect and contamination in an oil-based mud environment. The method aims to address quality concerns in challenging well profiles where conventional cementing techniques are deemed to fail, which potentially lead to additional cost for remedial cementing work and non-productive rig time. The method is utilized to tackle a re-occurring issue of tagging soft cement plugs that are used for several applications, including plugging and abandonment and sidetracking operations, in deep highly deviated or horizontal wells. In such profiles, fluids and cement don't balance themselves as gravity is not the predominant factor. Subsequently, the cement plug cannot be balanced, and contamination takes place resulting in its failure. The method targets minimizing this by reducing the effect of pulling out of the plug by keeping the drill pipe stinger inside the cement slurry and replacing the volume created while pulling out of the hole by pumping cement until completing the placement of the plug. The technique is time sensitive since most of the cement slurry remains in drillpipe while preforming the job. For that reason, intensive planning is required to ensure successful implementation. One factor that impact the quality of the job is the abundance of actual downhole data, including caliper and temperature logs. Based on that, cement slurry design can be adjusted to account for temperature effect as well as adjusting volume calculations. Also, lab and computer simulations play a significant role in determining several parameters such as spacer formulation, mud removal efficiency, ultimate compressive strength chart, and additives concentrations, especially retarders. In terms of field preparedness, specific equipment and competencies are required for such critical jobs. The method was deployed successfully in several jobs, resulting in excellent cement quality. At the top of that, the method was used to optimize operations by pumping a single long cement plug instead of several balanced cement plugs. This has directly resulted in saving several rig days while delivering satisfactory cement hardness. A couple of case studies are introduced to showcase the effectiveness of the method in challenging applications. Pump and pull technique proved that it resolves the challenge of placing cement plugs in highly deviated and horizontal wells. It earns more importance as wells are growing in complexity. Recently, the technique was used to successfully place significantly longer cement plugs, achieving an outstanding level of operational efficiency by saving rig days and resulting in cost optimization of cementing operations by reducing the number of required plugs to be pumped.
Enhanced oil recovery (EOR) projects are generally considered to be of high cost and risk. Modeling and prediction of performance remains a challenging task. Artificial Intelligence (AI) techniques have the potential to aid in performance optimization of EOR projects that results in increase in hydrocarbon recovery. The classical approaches to EOR optimization have either relied on the human expertise that utilizes low scale analysis, experience or physics-based reservoir simulation that requires long lead time to build. These are difficult to use to evaluate huge number of possible optimization scenarios. A combined approach of physics-based and data-driven AI models can provide a more efficient way for EOR performance optimization. This paper investigates the active research of deployment of AI in this particular application. Modeling of EOR performance optimization includes data assimilation, model fitting, validation, and testing. EOR performance optimization is investigated using a single- and multi- objective problem approach. The goal of creating the models is to provide a quantitative outcome, and for the model to be continuously updated to reflect actual field data and conditions. A case study of a steam flooding EOR project in San Joaquin Basin is presented. The project optimization included building combined physics-based and data-driven AI models to maximize production, maximize NPV, and minimize steam injection. Enormous number of possible injection plans were generated to select the best scenario with consideration of field and operator constrains. The modeling was used as a basis in a significant change in the field development strategy. Assessment after plan implementation showed a remarkable incremental gain in production in comparison to expected decline. The research reviews challenges of this type of modelling when it comes to data quality needs, long term implementations, and potential of rapid add up of computing power needs. Artificial intelligence offers opportunities to improve EOR performance optimization processes. It provides the capability of modeling to perform analysis and provide a fast and actionable plan as well as to predict production forecast. Also, it gives a powerful tool for day-to-day decision making, which is needed in this type of projects. The ever-evolving nature of artificial intelligence techniques hold more promise for improvement, as it can address some of the short comes experienced today.
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.