Formation Resistivity Prediction Using Decision Tree and Random Forest

Ibrahim, Ahmed Farid; Abdelaal, Ahmed; Elkatatny, Salaheldin

doi:10.1007/s13369-022-06900-8

Cited by 9 publications

(2 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Machine learning (ML) can be used to analyze large and complex datasets to improve decision-making and automate tasks in the industry. ML has been used in various applications in the oil and gas industry such as seismic surveys, well logs, drilling parameters, and production data to create detailed models of reservoirs [25][26][27][28][29] .…”

Section: Machine Learning Applicationsmentioning

confidence: 99%

Data-driven models to predict shale wettability for CO2 sequestration applications

Ibrahim

Elkatatny

2023

Sci Rep

Self Cite

View full text Add to dashboard Cite

The significance of CO2 wetting behavior in shale formations has been emphasized in various CO2 sequestration applications. Traditional laboratory experimental techniques used to assess shale wettability are complex and time-consuming. To overcome these limitations, the study proposes the use of machine learning (ML); artificial neural networks (ANN), support vector machines (SVM), and adaptive neuro-fuzzy inference systems (ANFIS) tools to estimate the contact angle, a key indicator of shale wettability, providing a more efficient alternative to conventional laboratory methods. A dataset comprising various shale samples under different conditions was collected to predict shale-water-CO2 wettability by considering shale properties, operating pressure and temperature, and brine salinity. Pearson’s correlation coefficient (R) was utilized to assess the linearity between the contact angle (CA) value and other input parameters. Initial data analysis showed that the elements affecting the shale wettability are primarily reliant on the pressure and temperature at which it operates, the total organic content (TOC), and the mineral composition of the rock. Between the different ML models, the artificial neural network (ANN) model performed the best, achieving a training R2 of 0.99, testing R2 of 0.98 and a validation R2 of 0.96, with an RMSE below 5. The adaptive neuro-fuzzy inference system (ANFIS) model also accurately predicted the contact angle, obtaining a training R2 of 0.99, testing R2 of 0.97 and a validation R2 of 0.95. Conversely, the support vector machine (SVM) model displayed signs of overfitting, as it achieved R2 values of 0.99 in the training dataset, which decreased to 0.94 in the testing dataset, and 0.88 in the validation dataset. To avoid rerunning the ML models, an empirical correlation was developed based on the optimized weights and biases obtained from the ANN model to predict contact angle values using input parameters and the validation data set revealed R2 of 0.96. The parametric study showed that, among the factors influencing shale wettability at a constant TOC, pressure had the most significant impact, and the dependency of the contact angle on pressure increased when TOC values were high.

show abstract

Section: Machine Learning Applicationsmentioning

confidence: 99%

Data-driven models to predict shale wettability for CO2 sequestration applications

Ibrahim

Elkatatny

2023

Sci Rep

Self Cite

View full text Add to dashboard Cite

show abstract

“…ML has different techniques, such as ANN, functional networks, support vector machines, ANFIS, RF, and DT, which show a good performance for prediction and classification . Different ML models were developed for various purposes, such as fracture pressure prediction, density log generation, resistivity prediction, − pore pressure gradient prediction while drilling, − formation lithology prediction, coal pay zones identification, rock geomechanical parameters prediction, stuck pipe prediction, equivalent circulation density prediction, , rheological properties prediction, , carbon dioxide solubility estimation in ionic liquids, compressibility factor estimation, ROP prediction, water saturation prediction, early kick detection and estimation in managed pressure drilling, and lithology classification . ML is widely used in the petroleum industry at large and drilling engineering in different areas, such as drilling fluids, drilling problems, and well control. , In this study, DTs and RF were employed to provide two predictive models for viscometer readings that can be used to estimate different rheological properties, such as PV, YP, apparent viscosity (AV), n, and K, using the existing equations in the literature.…”

Section: Introductionmentioning

confidence: 99%

Data-Driven Framework for Real-time Rheological Properties Prediction of Flat Rheology Synthetic Oil-Based Drilling Fluids

2023

Self Cite

View full text Add to dashboard Cite

Appropriate mud properties enhance drilling efficiency and decision quality to avoid incidents. The detailed mud properties are mainly measured in laboratories and are usually measured twice a day in the field and take a long time. This prevents real-time mud performance optimization and adversely affects proactive actions. As a result, it is critical to evaluate mud properties while drilling to capture mud flow dynamics. Unlike other mud properties, mud density (MD) and Marsh funnel viscosity (MFV) are frequently evaluated every 15−20 min in the field. The goal of this study is to predict the rheological properties of flat rheology synthetic oil-based mud (SOBM) in real time using machine learning (ML) techniques such as random forest (RF) and decision tree (DT). A proposed approach is followed to first predict the viscometer readings at 300 and 600 RPM (R 600 and R 300 ) and then estimate the other mud properties using the existing equations in the literature. A set of data contained MD, MFV, and viscometer readings (R 300 and R 600 ) for different samples from the same mud type. The mud samples were collected after going through a shale shaker. MD and MFV are measured by a mud balance and a Marsh funnel, respectively, while rheology is evaluated using a viscometer. The data were randomly split into training, testing, and validation data sets. The ML models' performance was evaluated through average absolute percentage error (AAPE) and correlation coefficient (R). The proposed models predicted the viscometer readings as a middle stage with a low AAPE that did not exceed 4.5% for both models. The suggested models forecasted the rheological properties with a good degree of accuracy, with an AAPE being less than 7% for most of the parameters. The proposed models can save costs and time since there is no need to include additional tools in the rig location. Furthermore, these models will significantly aid in avoiding serious problems and achieving better rig hydraulics and hole cleaning, which in turn will technically and economically enhance drilling operations.

show abstract

An Efficient Approach for Accident Severity Classification in Smart Transportation System

2022

View full text Add to dashboard Cite

Formation Resistivity Prediction Using Decision Tree and Random Forest

Cited by 9 publications

References 29 publications

Data-driven models to predict shale wettability for CO2 sequestration applications

Data-driven models to predict shale wettability for CO2 sequestration applications

Data-Driven Framework for Real-time Rheological Properties Prediction of Flat Rheology Synthetic Oil-Based Drilling Fluids

An Efficient Approach for Accident Severity Classification in Smart Transportation System

Contact Info

Product

Resources

About