Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
New Approach for Stress-Dependent Permeability and Porosity Response in the Bakken Formation
Ouadi, Habib (University of North Dakota) | Mellal, Ilyes (University of North Dakota) | Chemmakh, Abderraouf (University of North Dakota) | Djezzar, Sofiane (University of North Dakota) | Boualam, Aldjia (University of North Dakota) | Merzoug, Ahmed (University of North Dakota) | Laalam, Aimen (University of North Dakota) | Mouedden, Nadia (University of North Dakota) | Khetib, Youcef (University of North Dakota) | Rasouli, Vamegh (University of Wyoming)
Abstract During the reservoir depletion and injection operations, the net effective stress is disrupted due to pore pressure changes. As a result, the reservoir properties, mainly porosity and permeability, are influenced by the change in the stress behavior in the reservoir rock. Understanding the porosity and permeability stress-dependent alteration is crucial since it directly impacts the reservoir storage capacity and the production/injection capabilities. Conventionally, lab experiments are conducted to understand the stress dependency of porosity and permeability magnitudes. Two methods are usually used: the unsteady-state method (Core Measurement System, CMS-300) and the steady-state method (Core Measurement System, CPMS). The challenges with these experiments reside in the fact that they are expensive and time-consuming and may cause the destruction of the core samples due to the applied stresses. This study aims to investigate the effect of stress variations on porosity and permeability changes. These properties were measured on a total of 2150 core data from the three members of the unconventional Bakken formation (upper, middle, and lower), applying 35 different Net Confining Stress (NCS) values, ranging from 400psi to 5800psi. A correlation was formulated between permeability and the NCS to illustrate the stress dependency relationships. The Grey Wolf Optimization algorithm (GWO) was used to tune the correlation for the Bakken formation. Machine Learning methods were also applied for the porosity and permeability stress dependency response prediction, which are as follows: Linear Regression (LR), Random Forest Regression (RF), XGBoost Regression (XGB), and Artificial Neural Network (ANN). The results demonstrate that the porosity and the permeability decrease with the increase of the NCS and vice versa. The permeability is highly sensitive to the NCS changes compared to the porosity. The developed correlations showed a good fit with the data extracted from the laboratory experiments of the pilot well. For the data-driven models, the coefficient of correlation R-Score ranged from 91% to 93%. These models can be used to constrain the modeling work and reduce the uncertainties by introducing the effect of the net effective stress changes during reservoir depletion/injection on petrophysical properties.
- North America > United States > South Dakota (1.00)
- North America > United States > North Dakota (1.00)
- North America > United States > Montana (1.00)
- (3 more...)
- Geology > Geological Subdiscipline > Geomechanics (1.00)
- Geology > Rock Type > Sedimentary Rock > Clastic Rock (0.69)
Comparative Analysis Between Different Artificial Intelligence Based Models Optimized with Genetic Algorithms for the Prediction of Oilfield Cement Compressive Strength
Ouadi, H. (The University of North Dakota) | Laalam, A. (The University of North Dakota) | Chemmakh, A. (The University of North Dakota) | Merzoug, A. (The University of North Dakota) | Mouedden, N. (The University of North Dakota) | Rasouli, V. (The University of North Dakota) | Djezzar, S. (The University of North Dakota) | Boualam, A. (The University of North Dakota) | Hassan, A. (College of Petroleum Engineering & Geosciences, King Fahd University of Petroleum & Minerals)
ABSTRACT: A proper cementing job is an essential application in any successful drilling operation since it is heavily related to the well integrity. The cement quality performance used in this process is quantified by the cement compressive strength measured in three standard periods: 2, 7, and 28 days. The chemical-mineralogical synthesis of the cement and fineness factor significantly affects the cement strength for the wellโs life cycle to avoid unwanted fluid leakage. This study aims to develop two Artificial Intelligence algorithms: Artificial Neural Networks (ANN) and Support Vector Regression optimized by Genetic Algorithm (SVR-GA) to estimate the oilfield cement compressive strength in three periods based on the particle distribution, the size fraction and the chemical-mineralogical composition of the cement mixtures. The intelligent models are validated with 98 laboratory samples to investigate their prediction performances. The ANN displays a strong relationship with the experimental data with a 98.7%, 87.9% and 97.5% coefficient of correlation for 2, 7 and 28 days respectively. The SVR-GA exhibit a higher accuracy with 98%, 98% and 97.5% coefficient of correlation for 2, 7 and 28 days respectively. Our study demonstrates the accuracy of algorithm performance of the cement compressive strength prediction for better well integrity problems elimination. 1. INTRODUCTION Well cementing is an important operation in the oilfield well development; it consists of pumping the cement slurry between the casing steel and the formation represented by the annulus (Nelson and Guillot, 2006). This process is considered a solution to prevent fluid communication between different geological formations (for instance, salt formation and aquifers) and the leakage pathways created toward the surface (Costa et al., 2021). A good cementing operation will ensure adequate zonal isolation and well integrity related to the ability to control the upward undesirable fluid migration for a solid well life cycle (Iyer et al., 2022). Serval factors impact the cement sheath contact with the formation of rock from one side and the casing steel from the other side, increasing the probability of cement failure (Krakowiak et al., 2015). These factors include the fluid injection due to underground storage or water disposal, pressure testing, drilling vibrations, high pressure and temperature (HPHT), formation shifting due to the reservoir depletion or fault reactivation, salt creeping, casing shrinkage, chemical reactions with the formation fluids and minerals, hydraulic fracturing and perforating. The presence of each of the previously mentioned factors will cause the fissures creation within the cement sheath due to the compression, traction, or microannulus, which reduces the cement integrity performances (Phyoe et al., 2015). In fact, laboratory experiment testing is fulfilled before any oilfield cementing operation to study the cement slurry proprieties in different conditions. The cement concrete quality control is mandatory, which can be adjusted by proposing some additives and materials to enhance the cement proprieties and efficiency. These experiments are costly and time-consuming, for instance, the stability test, density test, thickening time, compressive strength free water test, fluid loss, and rheology test (Lenin Diaz, 2016). The compressive strength is the most critical mechanical propriety to secure a strong cement bond. It is a destructive test in which the cement simple is under axial loading conditions until it reaches the failure point. The direct measurement tool is called the uniaxial compressive test (Alkinani et al., 2021).
- North America > United States > North Dakota (0.93)
- Asia > Middle East (0.93)
- Africa > Middle East > Algeria > Ouargla Province (0.28)
- Geology > Mineral (0.69)
- Geology > Geological Subdiscipline > Mineralogy (0.54)
- Information Technology > Artificial Intelligence > Machine Learning > Neural Networks (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Statistical Learning > Support Vector Machines (0.88)
- Information Technology > Artificial Intelligence > Machine Learning > Statistical Learning > Regression (0.67)
Numerical Simulation of Gas Lift Optimization Using Artificial Intelligence for a Middle Eastern Oil Field
Al-Janabi, Mohammed Ahmed (Missan Oil Company) | Al-Fatlawi, Omar F. (University of Baghdad) | Sadiq, Dhifaf J. (University of Baghdad) | Mahmood, Haider Abdulmuhsin (Basrah Oil Company) | Al-Juboori, Mustafa Alaulddin (Iraqi Drilling Company)
Abstract Artificial lift techniques are a highly effective solution to aid the deterioration of the production especially for mature oil fields, gas lift is one of the oldest and most applied artificial lift methods especially for large oil fields, the gas that is required for injection is quite scarce and expensive resource, optimally allocating the injection rate in each well is a high importance task and not easily applicable. Conventional methods faced some major problems in solving this problem in a network with large number of wells, multi-constrains, multi-objectives, and limited amount of gas. This paper focuses on utilizing the Genetic Algorithm (GA) as a gas lift optimization algorithm to tackle the challenging task of optimally allocating the gas lift injection rate through numerical modeling and simulation studies to maximize the oil production of a Middle Eastern oil field with 20 production wells with limited amount of gas to be injected. The key objective of this study is to assess the performance of the wells of the field after applying gas lift as an artificial lift method and applying the genetic algorithm as an optimization algorithm while comparing the results of the network to the case of artificially lifted wells by utilizing ESP pumps to the network and to have a more accurate view on the practicability of applying the gas lift optimization technique. The comparison is based on different measures and sensitivity studies, reservoir pressure, and water cut sensitivity analysis are applied to allow the assessment of the performance of the wells in the network throughout the life of the field. To have a full and insight view an economic study and comparison was applied in this study to estimate the benefits of applying the gas lift method and the GA optimization technique while comparing the results to the case of the ESP pumps and the case of naturally flowing wells. The gas lift technique proved to have the ability to enhance the production of the oil field and the optimization process showed quite an enhancement in the task of maximizing the oil production rate while using the same amount of gas to be injected in the each well, the sensitivity analysis showed that the gas lift method is comparable to the other artificial lift method and it have an upper hand in handling the reservoir pressure reduction, and economically CAPEX of the gas lift were calculated to be able to assess the time to reach a profitable income by comparing the results of OPEX of gas lift the technique showed a profitable income higher than the cases of naturally flowing wells and the ESP pumps lifted wells. Additionally, the paper illustrated the genetic algorithm (GA) optimization model in a way that allowed it to be followed as a guide for the task of optimizing the gas injection rate for a network with a large number of wells and limited amount of gas to be injected.
- North America > United States (1.00)
- Asia > Middle East > Iraq > Diyala Governorate (0.60)
- Overview (1.00)
- Research Report > New Finding (0.86)
- South America > Brazil > Rio de Janeiro > South Atlantic Ocean > Campos Basin > Marlim Field > Macae Formation (0.99)
- South America > Brazil > Rio de Janeiro > South Atlantic Ocean > Campos Basin > Marlim Field > Lago Feia Formation (0.99)
- Europe > United Kingdom > North Sea > Northern North Sea > East Shetland Basin > Block 211/28 > North West Hutton Field > Brent Group Formation (0.99)
- (5 more...)
Optimizing Fishbone Multilateral Well Productivity Forecasting with Machine Learning Techniques
Silva, Henry Galvis (Texas A&M University, College Station) | Conde, Oliver Rojas (Texas A&M University, College Station) | Khalifa, Houdaifa (University of North Dakota, Grand Forks) | Benarbia, Achouak (University of North Dakota, Grand Forks) | Montes, Jose Carlos Cardenas (Ecopetrol)
Abstract Fishbone drilling (FbD), a premier method in multilateral well drilling, has been adopted in numerous global hydrocarbon fields. Its application leads to notable improvements in recovery rates while diminishing carbon emissions from the drilling process. The essence of FbD lies in constructing several minor holes that radiate in diverse directions. However, a significant obstacle with Fishbone drilling is the absence of comprehensive models to understand the impact of each Fishbone variable on total production, and their interplay. This research delves into the intricacies of fishbone well productivity prediction by leveraging a suite of machine learning algorithms, including RandomForest, GradientBoost, LinearSVC, AdaBoost, and KNeighbors. Through extensive model evaluation metrics such as MSE, RMSE, MAE, and the R-squared score, the study offers insights into the relative weightage of input features, with bottom-hole pressure emerging as paramount. The culmination of the research is the DrillFish Web App, an innovative platform designed for industry professionals to predict fluid flowrate and optimize fishbone drilling designs.
- North America > United States > North Dakota (0.46)
- North America > United States > New Mexico (0.29)
- North America > United States > Texas (0.28)
- Asia > Middle East > Saudi Arabia (0.28)
- North America > United States > South Dakota > Williston Basin > Bakken Shale Formation (0.99)
- North America > United States > North Dakota > Williston Basin > Bakken Shale Formation (0.99)
- North America > United States > Montana > Williston Basin > Bakken Shale Formation (0.99)
- North America > United States > Texas > Fort Worth Basin > Barnett Field > Barnett Shale Formation (0.98)
- Information Technology > Artificial Intelligence > Machine Learning > Neural Networks (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Ensemble Learning (0.72)
- Information Technology > Artificial Intelligence > Machine Learning > Statistical Learning > Nearest Neighbor Methods (0.31)
Statistical Analysis of the Petrophysical Properties of the Bakken Petroleum System
Laalam, Aimen (University of North Dakota) | Ouadi, Habib (University of North Dakota) | Merzoug, Ahmed (University of North Dakota) | Chemmakh, Abderraouf (University of North Dakota) | Boualam, Aldjia (University of North Dakota) | Djezzar, Sofiane (University of North Dakota) | Mellal, Ilyas (University of North Dakota) | Djoudi, Meriem (University of Boumerdes)
Abstract Porosity and permeability represent the main parameters for an accurate petrophysical evaluation. These parameters are often evaluated either from well logs interpretation or core data. The wireline logs provide continuous measurements of physical rock properties and can be interpreted to provide porosity and permeability indirectly. Thus, it has to be inferred through relationships with core data from the same field or well or from empirically derived equations. Another approach is to model the relationship between porosity and permeability from core data which provide more accurate estimations but are expansive and cannot be acquired at every depth on every well. In addition, machine learning technics gained a lot of importance in solving similar problems. To produce a continuous permeability from a computed porosity in any well, we use statistical analysis on core data to obtain a correlation between porosity and permeability for a particular formation. This paper aims to generate permeability-porosity data-driven models for the Bakken formation, representing an unconventional reservoir within the Williston Basin in the US, using 426 core data with a wide range of porosity and permeability. Different machine learning algorithms have been developed including Linear Regression (LR), Artificial Neural Network (ANN), Random Forest Regressor (RFR), Extreme Gradient Boosting (XGBoost), Adaptive Booster Regressor (AdaBoost), and Support Vector Regression (SVR), to predict the permeability from porosity. Evaluating the obtained correlation and the machine learning algorithms was based on the R score, the Minimum Squared Error (MSE), and the Mean Absolute Error (MAE) as evaluation metrics. The developed models yielded an R score ranging from 0.61 to 0.74, with the ANN model outperforming the other algorithms resulting in the highest R score and lowest error. The models were evaluated on unseen data from other wells drilled in the same formation, and a good match of permeability was obtained. Introduction Permeability is one of the important petrophysical parameters in reservoir engineering (Zhao et al., 2022). an accurate determination is vital for hydrocarbons recovery, carbon storage, geothermal energy capacity, and well placement optimization (Byrnes, 1994). It has a strong relationship with reservoir quality evaluation and the fluid production capacity (Baas et al., 2007), which is based on reservoir characterization, the flow unit identification, and the location of the perforation intervals (Doyen, 1988; Pittman, 1992). It represents the ability of the rock to transmit fluid (Ghafoori et al., 2008). The accurate permeability detection affects the economic value of hydrocarbons accumulation and petroleum reservoir management before and during production.
- North America > United States > North Dakota (1.00)
- North America > Canada > Saskatchewan (1.00)
- North America > Canada > Manitoba (1.00)
- Asia > Middle East > Israel > Mediterranean Sea (0.24)
- Geology > Rock Type > Sedimentary Rock > Clastic Rock (0.69)
- Geology > Geological Subdiscipline > Geomechanics (0.66)
- Geology > Geological Subdiscipline > Economic Geology > Petroleum Geology (0.50)
- North America > United States > South Dakota > Williston Basin > Bakken Shale Formation (0.99)
- North America > United States > North Dakota > Williston Basin > Bakken Shale Formation > Middle Bakken Shale Formation (0.99)
- North America > United States > Montana > Williston Basin > Bakken Shale Formation (0.99)
- (2 more...)
- Information Technology > Artificial Intelligence > Machine Learning > Statistical Learning > Regression (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Neural Networks (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Ensemble Learning (1.00)
- Information Technology > Artificial Intelligence > Machine Learning > Statistical Learning > Support Vector Machines (0.88)