Oil Field Data Research Articles

Determination the depth and thickness of the Yamama Formation using the reflection seismic method in the Nasiriyah field-Southern Iraq

An interpretive study of the Yamama Formation for seismic data in the Nasiriyah oil field was achieved in this work. We choice this location for our study because of its important location between several oil fields. Using SERVPRO-10, the NS-1 well data and the seismic data were represented by time, depth, velocity, thickness, and three-dimensional mapping of the thickness distribution. The seismic interpretation of the upper and lower-time maps of the Yamama Formation showed the presence of faults towards the southwest and northeast as a result of the influence of the transverse fault system. Structural features of the Yamama Formation was also identified from the depth maps, with seven closures identified in the upper boundary and three in the lower boundary, representing the lowest occurrence of structural influence. The 3D map showed the highest value of formation thickness in the northeast direction, and that represents the distribution of formation rock units the directions of increasing sediment thickness, and the deposition location from the shelf margin to the sedimentary basin, indicates the increasing in sedimentation energy of the basin. Petrophysical properties of the Yamama Formation maps show a significant increase in the value of density and velocity and its relationship with acoustic impedance and a decrease in porosity from northeast to southwest directions

Enhanced petrophysical evaluation through machine learning and well logging data in an Iranian oil field

Reservoir petrophysical assessments are essential for determining hydrocarbon reserves, production, and characterizing reservoir layers. Advanced logging technology identifies crucial petrophysical parameters, including porosity type, rock pore size and type, and static/dynamic properties. The aim of this study is to present a petrophysical evaluation of the studied reservoir and to identify the reservoir layers by calculating and determining petrophysical indicators using well logging data. Additionally, various machine learning methods, including Adaptive Neuro-Fuzzy Inference System, Extreme Learning Machine, Multi Gene Genetic Programming, Decision Tree, and Adaptive Boosting, were compared to model the water saturation data according to different logs. The investigated depth ranged from 4050.6 to 4560 m, with each image containing over 3000 data at the desired depth. The main lithology of the formation was limestone with some shale. By conducting a petrophysical evaluation and applying parameter cutoffs, productive zones within the reservoir were identified. Layer 3 had the highest average net porosity (18%) and net water saturation (17%), with secondary porosity observed in most layers. Among the machine learning models tested the AdaBoost model demonstrated the lowest error value for estimating water saturation, with an RMSE of 0.0152 and an AARE% of 3.1610, establishing it as the most effective model in this study. Furthermore, the GP model provided a correlation between the input parameters and predicted water saturation, demonstrating good accuracy with an RMSE of 0.0231 and an AARE of 4.3597.

Study on the Accumulation Model of the Cretaceous Reservoir in AHDEB Oilfield, Iraq

The Ahdeb oil field is located in the southwestern part of the Zagros fold deformation zone. The study of the model of the formation of the oil reservoir in this field will be helpful to deepen the pattern of hydrocarbon distribution in this zone. In this paper, we use the seismic data of the Ahdeb oil field to recover the tectonic evolution history of the field. Under neotectonic movement, the oil field formed in the early stage, migrated to the high point in the late stage, and finally entered the present formation. From here, for the oil-bearing inclusions within the reservoir, the photometric absorption values of the organic matter groups were measured by infrared spectroscopy. Their ratios were used to evaluate the maturity, thus discovering two phases of oil charging. Finally, using the hydrocarbon generation history and tectonic evolution history, combined with the oil and gas transportation periods in the reservoir, we deduce that the reservoir formation mode in the area is a two-phase gathering and final adjustment formation mode. This understanding of the hydrocarbon formation patterns will promote oil and gas exploration in this zone.

An Efficient Method for Identifying Inter-Well Connectivity Using AP Clustering and Graphical Lasso: Validation with Tracer Test Results

Identifying inter-well connectivity is crucial for optimizing reservoir development and facilitating informed adjustments. While current engineering methods are effective, they are often prohibitively expensive due to the complex nature of reservoir conditions. In contrast, methods that utilize historical production data to identify inter-well connectivity offer faster and more cost-effective alternatives. However, when faced with incomplete dynamic data—such as long-term shut-ins and data gaps—these methods may yield substantial errors in correlation results. To address this issue, we have developed an unsupervised machine learning algorithm that integrates sparse inverse covariance estimation with affinity propagation clustering to map and analyze dynamic oil field data. This methodology enables the extraction of inter-well topological structures, facilitating the automatic clustering of producers and the quantitative identification of connectivity between injectors and producers. To mitigate errors associated with sparse production data, our approach employs sparse inverse covariance estimation for preprocessing the production performance data of the wells. This preprocessing step enhances the robustness and accuracy of subsequent clustering and connectivity analyses. The algorithm’s stability and reliability were rigorously evaluated using long-term tracer test results from a test block in an actual reservoir, covering a span of over a decade. The results of the algorithm were compared with those of the tracer test to evaluate its accuracy, precision rate, recall rate, and correlation. The clustering results indicate that wells with similar characteristics and production systems are automatically grouped into distinct clusters, reflecting the underlying geological understanding. The algorithm successfully divided the test block into four macro-regions, consistent with geological interpretations. Furthermore, the algorithm effectively identified the inter-well connectivity between injectors and producers, with connectivity magnitudes aligning closely with actual tracer test data. Overall, the algorithm achieved a precision rate of 79.17%, a recall rate of 90.48%, and an accuracy of 91.07%. This congruence validates the algorithm’s effectiveness in the quantitative analysis of inter-well connectivity and demonstrates significant potential for enhancing the accuracy and efficiency of inter-well connectivity identification.

Transfer learning for well logging formation evaluation using similarity weights

Machine learning has been widely applied in well logging formation evaluation studies. However, several challenges negatively impacted the generalization capabilities of machine learning models in practical implementations, such as the mismatch of data domain between training and testing datasets, imbalances among sample categories, and inadequate representation of data model. These issues have led to substantial insufficient identification for reservoir and significant deviations in subsequent evaluations. To improve the transferability of machine learning models within limited sample sets, this study proposes a weight transfer learning framework based on the similarity of the labels. The similarity weighting method includes both hard weights and soft weights. By evaluating the similarity between test and training sets of logging data, the similarity results are used to estimate the weights of training samples, thereby optimizing the model learning process. We develop a double experts' network and a bidirectional gated neural network based on hierarchical attention and multi-head attention (BiGRU-MHSA) for well logs reconstruction and lithofacies classification tasks. Oil field data results for the shale strata in the Gulong area of the Songliao Basin of China indicate that the double experts’ network model performs well in curve reconstruction tasks. However, it may not be effective in lithofacies classification tasks, while BiGRU-MHSA performs well in that area. In the study of constructing large-scale well logging processing and formation interpretation models, it is maybe more beneficial by employing different expert models for combined evaluations. In addition, although the improvement is limited, hard or soft weighting methods is better than unweighted (i.e., average-weighted) in significantly different adjacent wells. The code and data are open and available for subsequent studies on other lithofacies layers.

Delineation of an inversion structure of Upper Jurassic-Lower Cretaceous succession based on interpretation of seismic data of Alamein oil field, Northern Western Desert, Egypt

Delineation of an inversion structure of Upper Jurassic-Lower Cretaceous succession based on interpretation of seismic data of Alamein oil field, Northern Western Desert, Egypt

Predicting Pump Inspection Cycles for Oil Wells Based on Stacking Ensemble Models

Beam pumping is currently the broadly used method for oil extraction worldwide. A pumpjack shutdown can be incurred by failures from the load, corrosion, work intensity, and downhole working environment. In this study, the duration of uninterrupted pumpjack operation is defined as the pump inspection cycle. Accurate prediction of the pump inspection cycle can extend the lifespan, reduce unexpected pump accidents, and significantly enhance the production efficiency of the pumpjack. To enhance the prediction performance, this study proposes an improved two-layer stacking ensemble model, which combines the power of the random forests, light gradient boosting machine, support vector regression, and Adaptive Boosting approaches, for predicting the pump inspection cycle. A big pump-related oilfield data set is used to demonstrate the proposed two-layer stacking ensemble model can significantly enhance the prediction quality of the pump inspection cycle.

A STATISTICAL ANALYSIS OF GEOLOGICAL AND ENGINEERING PREDICTORS OF OILFIELD PERFORMANCE RESPONSE: A CASE STUDY OF OILFIELDS ON THE UK CONTINENTAL SHELF

Oilfield production is controlled by a wide range of geological and engineering parameters, many of which are at least partially interrelated. This paper uses multivariate statistical methods (principal component analysis, regression analysis and analysis of variance) to determine how these parameters are related, and which of them are most significant in controlling and predicting oilfield performance. The analysis is based on a database of publicly available oilfield data from the United Kingdom Continental Shelf (UKCS), from which a series of geological, engineering and fluid‐related control variables from 136 fields were pre‐processed and analyzed. This dataset is a subset of a much wider project database for UKCS oil, gas and condensate fields. For this study, the project database was divided into two datasets: a first dataset with 10 parameters from 136 fields, and a second, more detailed dataset with 27 parameters from 38 fields. Both datasets were analysed using principal component analysis in order to investigate possible correlations between numerically/statistically interrogable predictor variables such as porosity, permeability, number of production wells, gas‐oil ratio and reservoir temperature. A regression analysis was then carried out on the predictor variables in order to obtain a ranking of predictability (i.e. how indicative a predictor is of a particular outcome) and sensitivity (how sensitive an outcome is to slight changes in a predictor) in relation to recovery factor based on R‐squared and regression coefficient values. The results showed that key variables from the principal component analysis included field size, number of production wells, PVT, gross depositional environment and reservoir quality. High‐ranking parameters of predictability and sensitivity from the regression analysis were found to include API, net‐to‐gross, porosity and reservoir depth. These results are consistent with previous studies and suggest that it should be possible to forecast oilfield recovery based on only a few selected input variables. As a preliminary test of forecasting ability of the variable permutations put forward, a best‐subsets multiple regression was carried out using a statistical software package and yielded results which corroborated the main findings.

Unconventional Wells Interference: Supervised Machine Learning for Detecting Fracture Hits

The primary objective of the study was development of a machine learning (ML)-based workflow for fracture hit (“frac hit”) detection and monitoring using shale oil-field data such as drilling surveys, production history (oil and produced water), pressure, and fracking start time and duration records. The ML method takes advantage of long short-term memory (LSTM) and multilayer perceptron (MLP) neural networks to identify the frac hits due to hydraulic communication between the fracking child well(s) and the producing parent well(s) within the same pad (intra-pad interaction) and/or on different pads (inter-pad interaction). It utilizes time series of pressure and production data from within a pad and from adjacent pads. The workflow can capture time variable features of frac hits when the model architecture is deep and wide enough, with enough trainable parameters for deep learning and feature extraction, as demonstrated in this paper by using training and testing subsets of the field data from selected neighboring pads with over a couple of hundred wells. The study was focused on frac-hit interaction among paired wells and demonstrated that the ML model, once trained, can predict the frac-hit probability.

Study on multi-factor casing damage prediction method based on machine learning

Casing damage is one of the common problems encountered in reservoir development, which seriously affects the normal production of the oil field. In this study, through the analysis of oil field data, a casing damage model under the coupling effects of mudstone hydration-corrosion and sand production-corrosion was established. Thirty-four influencing factors of casing damage were determined in four categories: geology, engineering, development, and corrosion. Six machine learning methods were used to predict the probability of casing damage under the coupling effects of multiple factors. The generalization performance of the model was evaluated using the recall rate of casing damage wells and accuracy. The results show that the random forest and LightGBM models show the best generalization performance. The prediction accuracy rates of the two models on the test set were 84.2% and 85.9%, respectively, and the random forest model showed an excellent performance of 92.3% on the recall rate of casing damage wells. Finally, the optimized model was used to perform sensitivity analysis on each influencing factor, and the main controlling factors of casing damage were obtained. Engineering measures to prevent casing damage are proposed. This study has made outstanding contributions to improving the economic benefits of the oilfield.

Numerical simulation of enhanced throughput of surfactant in Lamadian oilfield

Based on the actual data of Lamadian oilfield in terms of geological structure, reservoir fluid physical properties and production dynamics, a numerical simulation geological model of CMG STARS chemical flooding module is established, which is consistent with the static and dynamic characteristics of the oil reservoir. In order to study the effect of enhanced throughput of surfactants, based on the mechanism of enhanced throughput of surfactants, the numerical model was established by defining parameters such as oil-water interfacial tension at different concentrations, residual oil saturation, oil-water relative permeability curve, capillary pressure, isotonic point change and surfactant diffusion coefficient in the formation, etc. The numerical simulation of surfactant enhanced throughput was carried out to optimize the surfactant enhanced throughput parameters in Lamadian oilfield, and the sensitivity sequence of each injection and production parameter to the surfactant enhanced throughput and oil was determined. The final recommended injection and production parameters are surfactant injection amount (10000m3), injection concentration (0.5%) and brazing time (2d). It is predicted that the cumulative oil increase can reach 1310m3, the initial daily fluid production is controlled at 80m3/d, the daily oil production can be increased by 10m3/d, and the water cut can be reduced by 10%.

A data-driven workflow for prediction of fracturing parameters with machine learning

In the realm of unconventional reservoir hydraulic fracturing design, the conventional optimization of mechanistic model parameters is a time-consuming process that impedes its responsiveness to the swift demands of on-site development. This study, rooted in Xinjiang oilfield data, delves into the utilization of machine learning methods for extensive field data. The research systematically elucidates the training and optimization procedures of a production forecasting model, achieving effective optimization of hydraulic fracturing design parameters. By employing polynomial feature cross-construction generate composite features, feature filtering is performed using the maximal information coefficient. Subsequently, wrapper-style feature selection techniques, including ridge regression and decision trees, are applied to ascertain the optimal combinations of model input parameters. The integration of stacking during model training enhances performance, while stratified K-fold cross-validation is implemented to mitigate the risk of overfitting. The ultimate optimization of hydraulic fracturing design parameters is realized through a competitive learning particle swarm algorithm. Results indicate that the accuracy of the data-driven production forecasting model can reach 85%. This model proficiently learns patterns from mature blocks and effectively applies them to optimize new blocks. Furthermore, expert validation confirms that the optimization results align closely with actual field conditions.

Method for operative estimation of current reservoir pressure distribution based on the wells normal production data

In the article there has been proposed method of operative estimation of the current reservoir pressure distribution from normal well operation data. The method is based on the algorithm including current distribution calculation of function values for potential and filtration velocity in the oilfild selected area. The method allows to monitor the current distribution of the reservoir pressure in the productive formation of the area under consideration, as well as to assess the effectiveness of the impact on the reservoir in order to maintain the reservoir pressure. The implementation of the proposed method, demonstrated on the example of the "Neft Dashlary" (X horizon, block V) oilfield data, showed high accuracy of the obtained calculated values. The average value of the relative error of the calculated values of the reservoir pressure to the actual values of the downhole pressure measurements at the wells is not more than 1%, and the average calculated value of the reservoir pressure in the productive formation in the pilot area coincides with its actual present value. Keywords: reservoir; reservoir pressure; reservoir enhanced oil recovery; zonal impact; productive horizon; well productivity; diagnostics; filtration; monitoring; streamlines.

Using a Deep Neural Network with Small Datasets to Predict the Initial Production of Tight Oil Horizontal Wells

Due to its abundant reserves, tight oil has emerged as a significant substitute for conventional petroleum resources. It has become one of the focal points of exploration and research, and a new hot spot in global unconventional oil and gas exploration and development. This has led to a significant increase in the demand for forecasting the production capacity of tight oil horizontal wells. The deep neural network (DNN), as a mature model, has demonstrated significant advantages in many fields. However, due to the confidentiality and uniqueness of oilfield data, acquiring large datasets has become a challenge. Traditional methods using small datasets for training DNN models result in low accuracy and overfitting issues, which hinders the development of neural networks in the petroleum industry. This study aims to predict the initial production capacity of tight oil horizontal wells by using a small dataset of 650 data points through a DNN model. The research results indicate that pre-trained and fine-tuned DNNs outperform shallow neural networks, supporting vector machines, and DNN trained with traditional methods in terms of better generalization performance. Their accuracy reached 91.3%, demonstrating that it is reasonable to use a small dataset with pre-trained and fine-tuned DNN models.

Structural Interpretations for Zubair Formation in Kumait oil field, Using 3D Seismic Reflection Data, Southern-East Iraq

This study deals with the interpretation of structural 3D seismic reflection of the Kumait oil field in southern Iraq within the administrative boundaries of the Maysan Governorate. Synthetic seismograms are prepared by using available data of the Kt-1 oil field by using Petrel software to define and pick the reflector on the seismic section of the Zubair Formation, Which represents the Cretaceous Age. The study shows that the Kumait structure is an anticline fold. It is thought to be a structure trap caused by the collision of the Arabian and Iranian plates and trending in the same direction as driving factors in the area, which are from the northwest to the southeast, and the overall trend of strata is north and northeast. Seismic structure interpretation results show the top of the Zubair Formation and several structure closures. The stratification Phenomenon of the study area was clear evidence of the existence of closures which are probable to be hydrocarbon traps.

A New Approach for Production Prediction in Onshore and Offshore Tight Oil Reservoir

Rapid technological advances have accelerated offshore and onshore tight oil extraction to meet growing energy demand. Reliable tools to carry out production prediction are essential for development of unconventional reservoirs. The existed tri-linear analytical solutions are verified to be versatile enough to capture fundamental flow mechanisms and make accurate production predictions. However, these solutions are obtained in Laplace space with the Laplace transform and numerical inversion, which may lead to uncertainty in the solution. In this paper, a general analytical solution is derived in real-time space through integral transform and average pressure substitution. Namely, the partial differential equations describing subsurface fluid flow are firstly triple-integrated and then the obtained volume average pressure are replaced with the rate-dependent expressions. Furthermore, the ordinary differential equations related to oil rate are solved analytically in real-time space. To validate our model, this derived solution is verified against two numerical models constructed with two typical physical configurations. The great match indicates the accuracy and applicability of the analytical solution. According to the developed workflow, two field cases including offshore and onshore tight oilfield data are selected for history matching and production prediction. This new approach not only makes the obtained solution more simplified, but also helps field engineers diagnose flow patterns more quickly to better optimize production schemes.

Adaptive spatiotemporal neural networks based on machine learning for missing well-log prediction

The well log is the basis for understanding a geologic structure and evaluating petroleum reservoirs. It inevitably leads to missing logs due to borehole conditions and equipment failures. Existing machine-learning methods introduce convolutional neural networks (CNNs) in temporal networks to learn local morphological (spatial) features for improving prediction. However, they ignore the temporal background information of the logs and the differences in spatial features at different depth points. An adaptive spatiotemporal transformer (ASTT) is developed to effectively overcome these challenges, which consists of a temporal background encoder (TBE), a spatial encoder (SE), and a spatiotemporal decoder (STD). TBE introduces average pooling in the transformer encoder to learn the hidden geologic information in the extracted longitudinal temporal features. SE combines CNN and an attention mechanism to learn the spatial features of each depth point differentially. STD maps the extracted spatiotemporal features to the missing logs. Experimental results on real oilfield data indicate that our ASTT achieves excellent performance in terms of fitting degree and test error. The results in the cross-logs and crosswell cases demonstrate the generalization of ASTT.

Dynamic liquid level prediction in oil wells during oil extraction based on WOA-AM-LSTM-ANN model using dynamic and static information

Accurately predicting the dynamic liquid level is the key to energy efficient operation of oil wells, however, the dynamic liquid level of oil wells in the same area varies widely, and the existing methods cannot achieve uniform modeling of the dynamic liquid level of multiple wells with high accuracy. To this end, based on the parameters of the multi-wells production process, this paper proposes method for predicting the dynamic liquid level of oil wells based on a Long short-term memory (LSTM) with attention mechanism (AM) and artificial neural network (ANN) optimized by the whale optimization algorithm (WOA). First, the factors significantly related to the change in the dynamic liquid level are identified and divided into dynamic and static information. Dynamic features are extracted using AM-LSTM. AM can enhance the impact of important information when extracting dynamic features using LSTM; Static features are extracted through ANN; Finally, both dynamic and static features are used as inputs to ANN to predict the dynamic liquid level. Solve the prediction model parameter selection problem with WOA. Using historical oilfield data collected in the field for validation, the experiment proves that the proposed method in this paper is effective for predicting the dynamic liquid level of multi-wells. Therefore, this prediction model can be used as a tool to detect the dynamic liquid level, which can achieve the purpose of reducing energy consumption and improving efficiency during oil extraction.

Modeling and simulation of non-Newtonian fluid flows and heat transfer in a non-isothermal coiled tubing to oil well operations

The coiled tubing system has become an attractive option for oil industry operations in deep-water environments. Composed of a long, continuous, and flexible steel pipe, the coiled tubing provides a quick and safe intervention in active wells reducing the process time and cost. The fluid flow in curved pipes is characterized by centrifugal forces that significantly affect the frictional head loss and fluid heat transfer. The temperature variation affects the fluid physicochemical properties, such as rheology, density, thermal properties, and the cement setting time. To avoid the risk of the cement slurry setting before reaching the desired location, set retarders are added to its formulation. Therefore, it is essential to predict the temperature profile in coiled tubing to optimize the fluid formulation and process control, avoiding increased process time and cost. This research aimed to develop a mathematical model to simulate the heat transfer in coiled tubing fluid flow based on momentum, energy, and mass conservation. The mathematical model is composed of a set of partial differential equations spatially discretized using the finite volume technique. A case study was used, providing actual oilfield data to simulate the non-isothermic flow through the coiled tubing. Temperature data from a Brazilian offshore oil well abandonment process in two different operating conditions were used to validate the mathematical model. The results obtained in this paper indicated a good agreement between the model predictions and the oil well experimental data, with average relative errors of up to 2.5%.

New empirical scaling equations for oil recovery by free fall gravity drainage in naturally fractured reservoirs

Gas-oil gravity drainage is a recognized major contributor to production in fractured reservoirs. While various empirical and analytical methods have been proposed to model this process, many of them contain assumptions that are questionable or require parameters that are not accessible at the field level. The aim of this work is to provide new, easy-to-use scaling equations for estimating the recoverable oil through gravity drainage in naturally fractured reservoirs, considering the effects of resistance capillary pressure. To accomplish this, data from four oilfields undergoing gravity drainage, including rock properties (eight sets), block height (three sets), and fluid properties (four sets), were used to generate a wide range of recovery curves using a single porosity numerical simulation model. Aronofsky's and Lambert's functions were then utilized to match the generated recovery curves. Statistical analysis revealed that the Aronofsky's function is more accurate in replicating the recovery patterns, while the Lambert's function tends to overestimate the early-time oil recovery and underestimate the oil recovery at a later stage in the majority of cases. A sensitivity analysis was subsequently performed, revealing that parameters such as absolute permeability, viscosity of oil, height of block, gas and oil density, characteristics of relative permeability and capillary pressure curves and interfacial tension (IFT) influence the amount of time taken to achieve the final recovery. Of these parameters, absolute permeability has the most significant effect on the amount of time needed to attain the final recovery, while the effect of difference between oil and gas densities is the lowest. Consequently, two different expressions were developed using nonlinear multiple regression analysis of simulated gravity drainage data which can be combined with the Aronofsky model to substitute the rate convergence constant. The new scaling equations include the effects of capillary pressure and other relevant factors in gravity drainage simulations. Both forms show satisfactory accuracy, as evidenced by the statistical parameters obtained (R2 = 0.99 and MSE = 0.0019 for both established correlations). The new correlations were verified using a wide range of oilfield data and are expected to provide a better understanding of the recovery process in naturally fractured reservoirs.