Sand Production Prediction Research Articles

Prediction of Sand Production from Poorly Consolidated Formations Using Artificial Neural Network: A Case Study, Nahr Umr Formation in Subba Oil Field

Sand production in poorly consolidated formations is a significant challenge to the oil and gas industry, resulting in reduced production and compromised equipment integrity. The oil and gas industry has recently become interested in data analytics because it has significantly simplified, accelerated, and reduced the cost of data visualization. The intriguing new approaches in artificial intelligence development promise to offer long-term solutions to the increasing problems affecting the petroleum industry's operations. Any completion operation in sandstone formations is based on determining whether or not a well will produce sand. It is economically unfavorable to the Nahr Umr Formation to install sand control equipment for wells that don't produce sand. This study aims to create an efficient and precise artificial neural network that can predict sanding in sandstone formations. To achieve this, we developed a two-layered ANN with a back-propagation algorithm using JMP software. The model formulation included various geological and mechanical factors that could impact sand detachment, such as Young’s modulus (YME), Poisson’s ratio, minimum and maximum horizontal stresses (SHMIN and SHMAX), pore pressure, shale volume, and formation strength. The study acquired data from both local fields, specifically the Amarah oilfield, and non-local fields from the literature. Two sets were created from the data: a training set comprising 75% of the data, and a test set comprising 25% of the data, both of which were classified. The speed and accuracy of a learning process in an Artificial Neural Network depends on the size of the data set, the number of hidden layers, and the input parameters. The statistical measures such as accuracy, confusion matrix, root mean square error, and generalized RSquare demonstrate the strong efficacy of the created Artificial Neural Network model. The results indicate a significant likelihood of sand production occurring over the perforation intervals across Su-4 and Su-5 wells. The method's unique feature is its ability to isolate shale areas and identify weak areas capable of producing sand in sandstone formations. In conclusion, it is determined that an Artificial Neural Network employing the back-propagation algorithm is a simple, reliable, and easy-to-use method for accurately predicting sand production.

An Accurate Critical Total Drawdown Prediction Model for Sand Production: Adaptive Neuro-fuzzy Inference System (ANFIS) Technique

AbstractSand production causes many problems in the petroleum industry. The sand production is predicted to control it in the early stages. Therefore, accurate prediction of sand production has been considered substantial in achieving successful sand control. Critical total drawdown (CTD) can indicate the sand production. The main drawback of the previous studies in predicting CTD is their lack of accuracy. Thus, this study aims to develop an accurate CTD estimation prediction model employing a trend analysis and adaptive neuro-fuzzy inference system (ANFIS). The method is chosen because of its higher performance; the model is built based on 23 published datasets from the Adriatic Sea. The developed ANFIS model is evaluated using various methods, namely, trend analyses. Trend analyses are conducted to show the effects of the features on the CTD to present the physical behavior. The model’s performance was also evaluated using statistical error analyses. In addition, the ANFIS and previously published models were assessed. The trend analyses show the correct relationship between all features and the CTD. In addition, the trend analyses for the previous models are discussed. The results show that the proposed ANFIS method outperforms published methods with an R of 0.9984 and an absolute average percentage relative error (AAPRE) of 4.293%.

Sand Production Prediction with Machine Learning using Input Variables from Geological and Operational Conditions in the Karazhanbas Oilfield, Kazakhstan

AbstractThis paper describes a comprehensive approach to predict sand production in the Karazhanbas oilfield using machine learning (ML) techniques. By analyzing data from 2000 wells, the research uncovered the complex dynamics of sand production and emphasized the critical need for accurately predicting the peak sand mass and its occurrence time. ML techniques can have a significant impact on prediction of sand production and on the optimization of oilfield operation, which can be improved with the combined use of enriched training data and domain-specific knowledge. The research underscored the influence of geological factors, especially fault proximity, on prediction accuracy. Domain and field knowledge is needed to formulate different production scenarios for prediction purposes such that the relevant data can be selected for the training of ML models. Moreover, new metrics are needed to evaluate model performance as the applied method is tailored for different operational strategies. As the peak sand mass is considered a pivotal event in field operation, new metrics in terms of peak prediction accuracy and peak time prediction accuracy were introduced to evaluate the performance of ML models. A suite of ML algorithms was employed in the study, which demonstrated notable accuracy in the classification of sand-producing wells.

A comprehensive review of cased hole sand control optimization techniques: Theoretical and practical perspectives

Cased hole sand control is a critical aspect of oil and gas well management, ensuring efficient production by mitigating sand production issues. This review presents a comprehensive review of cased hole sand control optimization techniques, offering theoretical insights and practical perspectives to enhance well productivity and longevity. The review encompasses various aspects, including the challenges of sand production, the importance of sand control, and the evolution of sand control techniques. Sand production in oil and gas wells can lead to equipment damage, wellbore instability, and reduced production efficiency. Effective sand control is thus essential to maintain well integrity and maximize hydrocarbon recovery. This review explores the fundamental principles of sand control, highlighting the key parameters influencing sand production, such as reservoir properties, fluid properties, and wellbore geometry. The review discusses traditional sand control methods, such as gravel packing and standalone screens, along with emerging technologies like expandable sand screens and autonomous inflow control devices. It evaluates the advantages and limitations of each technique, considering factors such as installation complexity, cost-effectiveness, and long-term reliability. Furthermore, the review delves into advanced sand control optimization strategies, including the use of modeling and simulation tools to predict sand production and optimize sand control design. It discusses the integration of data analytics and machine learning techniques for real-time monitoring and decision-making, enhancing sand control effectiveness and reducing operational risks. Practical case studies and field applications are presented to illustrate the implementation of sand control optimization techniques in real-world scenarios. These case studies highlight the challenges faced, the solutions implemented, and the outcomes achieved, providing valuable insights for industry practitioners and researchers. In conclusion, this review provides a comprehensive overview of cased hole sand control optimization techniques, emphasizing the importance of integrating theoretical knowledge with practical considerations to enhance sand control efficiency and maximize well productivity. his review paper will systematically analyze existing methodologies and the latest advancements in cased hole sand control treatments. It explores theoretical frameworks, compares practical outcomes, and discusses the implications of optimized frac geometries for reservoir performance enhancement across varied geological settings.

Sand control during gas production from marine hydrate reservoirs by using microbial-induced carbonate precipitation technology: A feasibility study

Sand production is one of the bottleneck problems restricting the safe and efficient production of marine hydrate reservoirs. Current research on this issue mainly focuses on the active sand control methods such as screen mesh and gravel packing, however, passive methods are rarely reported. In this study, a novel sand control strategy which uses microbial-induced carbonate precipitation (MICP) technology to reinforce hydrate reservoirs around the well is proposed. Based on the geological data of hydrate reservoirs in Japan, a reservoir-scale microbial reaction-migration model is established. The feasibility of MICP reinforcement to prevent sand production is demonstrated by the combinations of indoor triaxial testing, productivity prediction and microscopic sand production prediction. The results show that the MICP technology can effectively enhance reservoir strength, and has a negligible impact on the gas recovery. Although the calcium carbonate generated will slightly decrease the porosity and permeability, the cumulative gas production does not significantly decrease and the water extraction is also prevented. Most importantly, the amount of sand production can sharply reduce because of the cementation of calcium carbonate among sand particles. Additionally, the effectiveness and applicability can be improved by selecting appropriate bacteria fixation methods and grouting pressure at different stages.

Predicting Onset of Sand Production in Oil Wells using Machine Learning

Sand production in oil wells is a significant challenge that negatively impacts productivity and compromise equipment integrity. This study explores the application of Optimized Support Vector Machine (SVM) binary classification algorithm to predict the onset of sand production in oil wells. A dataset from 63 oil wells was utilized, and class labels were determined based on the bulk and shear modulus product. The model development incorporated geological and mechanical parameters that could influence sand detachment such as: Young’s modulus, Poisson’s ratio, minimum and maximum horizontal stresses, overburden pressure, pore pressure, depth, fracture gradient, and formation strength. Instances above the threshold of 8E+11 were classified as indicative of no sand production, while those below were considered potential sand production scenarios. The SVM model demonstrated remarkable accuracy in predicting sand production onset, trained and tested rigorously with field data. The model's accuracy was evaluated using statistical parameters, such as: accuracy (ACC), sensitivity (SE), specificity (SP), and Matthew's Correlation Coefficient (MCC). From the results, the model achieved a score of 1 across all parameters, indicating high reliability and accuracy in sand production prediction. The practical implications of this model are significant, offering assistance to completion engineers in making proactive decisions regarding sand control strategies. Furthermore, the integration of this model into oil and gas industry processes can optimize operational efficiency by foreseeing potential sand production events, hence, preventing production impairment and ensuring loss prevention.

Prediction of wellbore sand production potential from analysis of petrophysical data coupled with field stress: a case study from the Shah-Deniz gas field (Caspian Sea Basin)

Identifying the optimal azimuth and inclination for wellbore drilling in sandy formations can be considered a valuable aid in reducing sand production risks, lost time, and decreasing drilling costs in the petroleum industry. Therefore, a numerical systematic approach was provided to predict sand production in wellbore SDX-5, drilled in a deep-water sandstone reservoir in the Shah-Deniz gas field (South Caspian Basin), which has never been done previously. Additionally, this systematic approach uses geomechanical and geodynamical criteria, along with petrophysical information (density and sonic log) and tectonic characteristics of the study area, which are influenced by the active tectonic stresses of the Apsheron-Balkhan zone. The subsurface data sources employed are more eco-friendly, available, and continuous than experimental tests. The computations conducted achieved azimuth, inclination, polar, and depth profile plots for the Lower Balakhany Formation. The calculations reveal that the optimum azimuth for the wellbore drilling trajectories is parallel to SHmax and oblique drilling to near horizontal is the result of optimum inclination. Polar plots showed optimum azimuth, inclination, and effect of wellbore trajectory on critical collapse pressure and collapse drawdown pressure with pressure values simultaneously, which identify safer alternatives for achieving higher petroleum production rates without sanding. Depth profile plots provide a simultaneous overview of the values of critical collapse pressure, critical sanding pressure for instantaneous drawdown, and optimum wellbore production pressure during drilling and production operations. Moreover, optimum reservoir fluid production (maximum discharge) rates can be identified and imposed as upper limits to prevent sand production.

Visualization-based experimental investigation of microscopic sand production characteristics and mechanisms in weakly consolidated sandstone reservoirs

Sand production is a troublesome problem in the petroleum industry, especially in relation to weakly consolidated formations. In this work, a visual experimental apparatus, which allows sand particles to be observed, has been constructed to simulate sand production in weakly consolidated sandstone core. The experimental apparatus can accommodate sandstone core flakes as a reactor. The experiment aims to investigate the microscopic sand production mechanism of the weakly consolidated sandstone reservoir under the influence of high water cut, high clay content and polymer aging blockage. Core flakes were produced using simulated formation sands with median particle sizes of 160 μm and uniformity coefficients of 2.5, 4.6, 7.3 and 12.6, as well as clay (content 5%, 10%, 20%, 30%, respectively) and aged polymers in the wells. A total of 17 sets of experimental tests were carried out with water injection. The visual experimental results show that sand morphology presents complex wormhole. Morphology is particularly affected by core strength and heterogeneity. Based on dynamic differential pressure of experimental results, most sand production process is divided into two obvious stages, that is, the cavities formation stage of sharp sand production and the stable sand production stage. Analysis of sensitive factors revealed that the sand production rate is strongly influenced by the core strength and related to clay content, formation heterogeneity and production program. The volumetric sand production ratio decreases with the increase of core strength, with the volumetric sand production ratio (VSPR) reaching up to 0.24%. It is found that about 38% clay (1–3.91 μm) was produced in the initial stage of low flow, which is particularly evident in cores with high clay content. Moreover, for weakly consolidated reservoir, the particle competition peeling mechanism of wormhole front is proposed, and the strengthening bonding effect of aging polymer and the pore blockage effect of fine clay are clarified. The findings could contribute to better understanding of the mechanisms underlying sand production in weakly consolidated sandstone reservoirs, which also provide a solid evidence base for accurate sand production prediction and more targeted sand control.

On the stress dependency of sand production coefficient in hydro-dynamical sanding criteria

Solids production is a complex physical process which is controlled by several factors including mechanical failure from in-situ stresses and hydrodynamic erosion from fluid flow. Hydrodynamic models for the prediction of sand production involve sanding criteria based on filtration theories. Such models contain a constitutive model parameter, the coefficient λ, with dimensions of inverse length which is calibrated by sand erosion tests, but its nature and its dependencies have not been clarified to date. The aim of this work is an attempt to refine the hydrodynamic models by investigating the dependence of the sand production coefficient λ on the external stress conditions and on the plastic zone Λ that is developed on hollow cylinders tests and propose an expression describing its importance in the sand production prediction modelling. The aim of the work is obtained through simulations with finite elements by utilizing the well-established Arbitrary Lagrangian-Eulerian (ALE) analysis considering the poro-mechanical coupling of the fluid-solid system for simulating hollow cylinder tests. Best fitting experimental data estimated values for the sand production coefficient for various values of external stress are obtained through back analysis. The dependence of λ on the external stress turns out to be fairly smooth and a three-parameter model is proposed to describe that dependence: a scale parameter, an exponent, and a stress parameter defining the magnitude of stress at which erosion onset is predicted. It turns out that the stress parameter is associated with the minimum stress required for plastic yielding to occur, which was also estimated theoretically. This finding is in agreement with the physical assumption underlying the simulations that erosion onsets and progresses after the material reaches a critical plastic strain as a consequence of material plastic yielding. A power law model describing the dependence of λ on the plastic zone depth is also proposed and discussed.

Experimental simulation of wormhole sanding cavity pattern and microscopic mechanism in heterogeneous weakly-cemented sandstone

Sand production has been a shared problem in the development of weakly-cemented sandstone oil reservoirs. Sanding simulation and prediction are of utmost importance for the production optimization of this type of reservoir. For a long time, research on sand production has been centered on “what is produced from the formation,” such as the size and rate of produced sand. However, “what is left inside the formation,” which is the structural change of the rock after sanding, is also another intriguing and important topic for the management of sand-prone reservoirs. Some related studies have been carried out, and they have proposed that wormhole-like pore throat will appear after sand production, but the precise morphological description and formation mechanism are still lacking. A series of sanding simulation experiments are performed to deepen the understanding of the sanding cavity pattern and its mechanism. The experiments are carried out using a visual sanding simulation apparatus. Through this, the complex wormhole sand production patterns are found and classified into single-branch wormhole cavity patterns and multi-branch wormhole cavity patterns. The extension processes of those different patterns are also demonstrated. Besides, this work discusses the change in the reservoir flowability performance in wormhole sanding mode, and the near-well flowability might be improved by actively inducing weakly-cemented sandstone to create a bigger aperture wormhole sanding pattern. Through the visual microscopic system, the sand competitive detachment mechanism that induces wormhole extending is revealed, along with the cavities concurrent extension mechanism that induces multi-branch wormhole extending. Moreover, this work discusses the microscopic detachment forms which help explain the sand-produced rate from weakly-cemented sandstone. This work enhances and creates a novel understanding of the sanding patterns and mechanisms in weakly-cemented heterogeneous reservoirs, which is beneficial to providing direct guidance for sand production prediction and sand control optimization.

Experimental study on sand production and coupling response of silty hydrate reservoir with different contents of fine clay during depressurization

To further understand the characteristics of clay and sand production (hereafter collectively referred to as sand production) and to provide optimization designs of sand control schemes are critical for gas production from clayey silt natural gas hydrate reservoirs in the South China Sea. Thus-, gas-water-sand production behavoirs and coupling reservoir subsidence characteristics before, during, and after hydrate dissociation of the clayey silt hydrate reservoirs with different clay contents (5%, 10%, 15%, 20%, 25%, and 30%) have been studied through a self-developed experimental system. The results show that with the increase of clay content, the total mass of sand production first increases and then decreases, and it reaches maximum when the clayey content is 20%. The sand production is the lowest before hydrate dissociation and increases significantly during hydrate dissociation, which mainly occurs in the high-speed gas and water production stage at the beginning of hydrate dissociation. After hydrate dissociation, the sand production decreases significantly. During the whole depressurization process, the clay and free sand particles generally move to the sand outlet due to the fluid driving force and overlying stress extrusion. However, for conditions of high clay contents, those particles fail to pass through the sand control screen and gradually accumulate and block the screen by forming a mud cake, which greatly reduce the permeability of the screen and limite sand production as well as gas and water production. Our research lays a foundation for sand production prediction and sand control scheme selection during gas recovery from clayey silty hydrate reservoirs that greatly need to consider a balance between sand control and gas productivity.

Experimental study on sand production of depleted oil and gas reservoirs

With the increasing demand for natural gas, the emergency peak shaving capacity of gas storage is facing great challenges. The operation of gas storage converted from depleted oil and gas reservoir has the characteristics of strong injection and production, and alternate injection and production. However, the sand production of the reservoir makes the major gas storages face the dilemma that the injection and production capacity cannot be enhanced. In this study, the impact factors of sand production in injection-production wells of gas storage converted from depleted oil and gas reservoir were analyzed, and the sand production simulation device of gas storage was developed. Aiming at the two main controlling factors of alternating load and flow, the experimental study of quantitative sand production was conducted. The experimental results show that the alternating load has a significant weakening effect on the reservoir, and the sand production is positively correlated with the number, size and span of the alternating load exerted on the reservoir. By the method of multiple regression analysis, the parameters affecting the sand production degree of the core were obtained as follows: the upper and lower limit pressure of the alternating load, the flow rate and the number of alternating loads. The sand production map was obtained by combining the sand production amount of the core and the upper limit pressure of the alternating load to evaluate the sand production degree of the injection-production wells. In addition, at the same flow rate, the ratio of sand yield of two alternating loads stages is approximately equal to the constant, which provides a reference for the prediction of sand production in gas storage. The indoor simulation experiment of sand production in gas storage provides technical ideas and theoretical basis for the reconstruction of gas storage in Nanpu 1–29 block, and guides the selection of sand control completion methods in gas storage.

Research advance of numerical simulation methods for sand production prediction of unconsolidated sandstone

Research advance of numerical simulation methods for sand production prediction of unconsolidated sandstone

Sand Production Prediction and Safe Differential Pressure Determination in a Deepwater Gas Field

Sand production is a critical issue during the development of offshore oil and gas fields. Certain gas fields (e.g. the AB gas field) have high porosity and high permeability, and with water at the bottom of the reservoir, the risk of sand production greatly increases at high differential pressures. Based on reservoir properties, geological conditions, production requirements, and well logging data, in this study an ultrasonic time difference method, a B index method, and a S index method are used together with a model of rock mass failure (accounting for water influx and pressure depletion) to qualitatively predict sand production. The results show that considered sample gas field has an overall high risk of sand production. The critical differential pressure (CDP) without water influx is in the range of 1.40 to 2.35 MPa, the CDP after water influx is from 0.60 to 1.41MPa. The CDP under pressure depletion is in the range of 1.20 to 1.92 MPa. The differential pressure charts of sand production are plotted, and the safe differential pressure windows with or without water influx are obtained. The model calculation results and the experimental results are consistent with the field production data, which indicates that the implemented prediction method could be taken as a reference for sand production prediction in similar deep water gas fields.

Prediction of Sonic Shear Wave Using Artificial Neural Network

Shear wave time is an important parameter participating in the calculation of rock mechanics properties. Many evaluation processes including wellbore stability analysis and sand production prediction are based on rock mechanics properties so these processes are directly related to the estimation of shear wave time. Several empirical correlations have been developed to predict shear wave time using regression analysis and artificial neural network techniques for the estimation of relationships between a dependent variable and one or more independent variables for certain conditions of the reservoir. However, they are not appropriate for reservoirs with different conditions as well as all effective parameters are not considered in previous relationships. In this study, the artificial neural network is adopted for predicting shear wave time using datasets consisting of 1922 data points for a certain directional oil well from Iraqi Fauqi oil field wells. Two sets of input parameters are tried: the first trial includes the readings of seven logs (Gamma-ray, caliper, compressional sonic wave, density, neutron, deep resistivity, true vertical depth), while the second trial includes the azimuth and the inclination angles in addition to the above seven readings. The optimum structure for both datasets is obtained using 12 neurons in a single hidden layer (ANN-7-12-1 and ANN-9-12-1). The statistical results reveal that an improvement is achieved when the well azimuth and inclination are included in the ANN model. A mathematical model with high performance using an artificial neural network has been developed. The mean square error and the determination coefficient for the developed model were 14.22 and 0.952 for ANN-7-12-1, while they were 9.62 and 0.966 for ANN-9-12-1, respectively. This study presents a simple mathematical model for further determination of shear wave velocities using ANN techniques which can be then integrated with the existent petroleum software programs.

Integrated Completion Design Method Predicts Sand Production

_ This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 205935, “From Wellbore Breakout to Sand-Production Prediction: An Integrated Sand-Control Completion Design Methodology and Case Study,” by Caleb DeValve, Gilbert Kao, SPE, and Stephen Morgan, ExxonMobil, et al. The paper has not been peer reviewed. _ This paper presents a physics-based approach to predicting sand production on a well-by-well basis to inform overall sand-management design. The work flow integrates geomechanical estimates of wellbore breakout and volume of failed sand downhole, log-based prediction of the sand-particle-size variation along the well path, modeling of sand filtration based on experimental and analytical methods for specific completion options, and a natural sandpack-permeability prediction for standalone screen (SAS) completions and associated well-performance analysis. Method The general challenge addressed in this paper is how to predict the produced sand mass rate from any given well with a specific sand-control completion installed. A novel four-step method has been created to predict sand production from wells drilled in hydrocarbon-bearing sandstone formations with installed sand-control equipment and its associated effect on well productivity (Fig. 1). This process improves the traditional sand-control completion design process, which relies on limited formation data obtained through coring programs and classified according to particle-size distribution (PSD) characteristics (D10, fines content, and uniformity coefficient). PSD is critical data for traditional sand-control analysis, and multiple cores are used to represent the varying rock characteristics expected throughout a reservoir. The selection of sand control and the ultimate completion design using existing work flows will be based on the range of PSDs measured from the core study, along with their predicted distribution along the drilled well paths based on practitioner experience. The novel work flow presented in this paper enhances the traditional approach by integrating geomechanical failed sand volume (FSV) predictions with sand-screen-filtration performance expectations, which are analyzed along the wellbore to improve overall completion selection. These results can be integrated with traditional well-performance simulation to understand completion performance and influence well startup.

Predicting Sanding Potential Using Empirical Method in “Ebendo” Field, Niger Delta, Nigeria

Sand formations represent a large fraction of oil and/or gas reservoirs in the world, hence it becomes imperative to determine the competency of the reservoirs to produce sand-free hydrocarbon. This informed the empirical technique adopted in this study to evaluate elastic parameters such as shear modulus (G), bulk compressibility (Cb), shear modulus to bulk compressibility (G/Cb) ratio and Unconfined Compressive Strength (UCS) to determine sand influx in the “Ebendo” Field, Niger Delta, Nigeria. To achieve this goal, seven (7) hydrocarbon-bearing sand units with thicknesses ranging from 12.51 to 48.63 m were identified at depth range of 1884.79 - 3350.15 m across the four (4) wells. These elastic parameters were estimated at the interval of interest. The range of values obtained for G/Cb ratio in EBD (01 and 02) is 1.49 × 1012 - 5.40 × 1012 psi2 while the range of values for G/Cb ratio in EBD (04 and 06) is 0.06 × 1012 - 0.41 × 1012 psi2. This result suggests that EBD (01 and 02) have no potential to sanding while EBD (04 and 06) have a high probability of sanding when compared to the threshold value of 0.8 × 1012 psi2. The production history of the “Ebendo” Field also correlates with the findings of this study. The low values of UCS in EBD (04 and 06) also agrees with the observation. Thus, this study has shown the efficacy of using empirical method as a quick approach to predicting sand production in the “Ebendo” Field and this technique could be used in other fields with similar geological setting.

Comparison of Machine Learning Algorithms for Sand Production Prediction: An Example for a Gas-Hydrate-Bearing Sand Case

This paper demonstrates the applicability of machine learning algorithms in sand production problems with natural gas hydrate (NGH)-bearing sands, which have been regarded as a grave concern for commercialization. The sanding problem hinders the commercial exploration of NGH reservoirs. The common sand production prediction methods need assumptions for complicated mathematical derivations. The main contribution of this paper was to introduce machine learning into the prediction sand production by using data from laboratory experiments. Four main machine learning algorithms were selected, namely, K-Nearest Neighbor, Support Vector Regression, Boosting Tree, and Multi-Layer Perceptron. Training datasets for machine learning were collected from a sand production experiment. The experiment considered both the geological parameters and the sand control effect. The machine learning algorithms were mainly evaluated according to their mean absolute error and coefficient of determination. The evaluation results showed that the most accurate results under the given conditions were from the Boosting Tree algorithm, while the K-Nearest Neighbor had the worst prediction performance. Considering an ensemble prediction model, the Support Vector Regression and Multi-Layer Perceptron could also be applied for the prediction of sand production. The tuning process revealed that the Gaussian kernel was the proper kernel function for improving the prediction performance of SVR. In addition, the best parameters for both the Boosting Tree and Multi-Layer Perceptron were recommended for the accurate prediction of sand production. This paper also involved one case study to compare the prediction results of the machine learning models and classic numerical simulation, which showed the capability of machine learning of accurately predicting sand production, especially under stable pressure conditions.

Investigating unary, binary and ternary interactive effects on pressure drawdown for optimal sand production prediction

Formation and oil well sand production is some of the greatest and oldest concerns for the Oil and Gas industry. The production of sand particles may vary from very small and limited amounts to far elevated levels which have the potential to block or plug the pore spaces near the perforated points to blocking production from surface facilities. Therefore, the timely and reliable investigation of conditions leading to the onset or quantification of sand production during production is imperative. The challenges of sand production are even more elevated while producing in reservoirs characterized by varying independent factors viscosity, flow rate, cementation, clay and fine sand content. Moreover, the combined interactive effects arising from varying these factors present more complexity during production since the combination could better depict actual field conditions. Investigating unary interactive effects discretely, in essence, may misrepresent actual field conditions since all factors do not act in isolation of others. This experimental research is based on the Design of Experiments (DOE) methodology to analyse factorial designs investigating the unary, binary and ternary interactive effects on sand production. An optimized regression model is developed to predict the pressure drawdown at sand production. The obtained results indicate that reservoirs characterized by varying (high and low) levels of viscosity, flow rate, cementation, clay and fine sand content have a resulting impact on sand production. Effects arising from unary interactions, i.e., flow rate and fine sand produce the highest significance levels compared to higher-order interactions, i.e., binary and ternary. The proposed regression model could be applied to predict the pressure drawdown as a precursor for sand production under viscous conditions.

Implications of machine learning on geomechanical characterization and sand management: a case study from Hilal field, Gulf of Suez, Egypt

Sand production is one of the major challenges in the oil and gas industry, so a comprehensive geomechanical analysis is necessary to mitigate sand production in mature fields. As the pore pressure drastically decline in depleted reservoirs, the sand production risk becomes more critical and needs to be studied. However, the absence of key logs in many wells is a big challenge in the petroleum industry, and most geologists and engineers use empirical equations to predict missed log intervals. We conducted a comprehensive geomechanical modeling study on a full set of logs from two wells from the Hilal field, Gulf of Suez, Egypt, to infer the geomechanical elements and predict sand production. We have used the multi-arm calipers to calculate the actual depth of damage ratio to validate the geomechanical parameters in the prognosis model and confirm the stress orientations. We used machine learning approach to infer key sonic log in X-10 well to replace the empirical equations. The multi-arm calipers analysis showed an observed anisotropy in the hole diameter size with more enlargement in the ENE direction and fits with the minimum horizontal stress direction in the direction of N 60oE. The later also deduced the maximum horizontal stress direction in N150 ° based on the induced fractures from borehole image data in a nearby field. We developed and compared two sand management models: one using empirical equation and the other using machine learning. The model driven by the Gardner equation suggests sand production from day one, which is not matched with the production data, while the model driven by machine learning suggests no sand production risk, which is matched with the actual production data. Our results demonstrate the advantage of using machine learning technique in geomechanical studies on the classical empirical equations in the area of study that can be applied in other basins. The findings of this study can help with a better understanding of the implications of machine learning on geomechanical characterization and sand management.