Prediction Of Reservoir Properties Research Articles

Prediction of reservoir properties and fluids using rock physics techniques in the Rio Del Rey Basin, Cameroon

The application of rock physics modeling has been a breakthrough in oil and gas industry and has greatly managed exploration and interpretation risk. In the Rio del Rey Basin, the rock physics method was used to ascertain the reservoir's characteristics and relate them to the elastic characteristics. Using conventional techniques to connect to the elastic properties and assess various rock physics diagnostics, the estimated petrophysical parameters were water saturations, porosity, and shale volume. By connecting the saturated bulk modulus to its porosity, the bulk modulus of the porous frame, the bulk modulus of the mineral matrix, and the bulk modulus of the pore filling fluids, the Gassmann fluid replacement modeling was used to simulate various fluid scenarios. Finally, an evaluation was conducted on rock physics boundaries and cement models. The reservoirs are rated as extremely good based on the results. The rock physics model generated showed the friable sand model is a suitable model for the said basin and the rock physics bounds that were evaluated indicated that the sediments are deposited below the critical porosity and less compacted thereby depicting a soft sand model. Conventional petrophysical study would not be able to reveal the clustering of the gas sands from the brine sand and the shale zone in connection to the geologic trend as demonstrated by the rock physics template. The anticipated model will be useful to extrapolate the study to other sedimentary basins in Cameroon and estimate porosity from the impedance acquired from seismic data throughout the basin.

Improved reservoir quality assessment by evaluating illite grain coatings, quartz cementation, and compaction – Case study from the Buntsandstein, Upper Rhine Graben, Germany

Reservoir quality (RQ) in the Buntsandstein of the Upper Rhine Graben is in the center of attention as it is a possible target formation for geothermal energy production and hydrocarbon exploration surrounding existing fields. An understanding of properties affecting reservoir quality of the target lithology is still fairly poor and the success of accurately targeting high RQ intervals is limited. This is due to the fact, that the effect of compaction and the interplay between enhanced chemical compaction (i.e. pressure dissolution of quartz grains), illitic contact coatings and quartz cementation in the lithology has been underestimated. The understanding and quantification of controlling factors on reservoir qualities in fluvio-eolian sedimentary rocks, deposited in a (semi-)arid climate has been improved in recent years and this case study highlights the benefits of detailed petrographic analyses in understanding diagenetic and compactional processes in this lithology. This is especially relevant in the observation of grain coating clay minerals, whose effect depends on their specific location, i.e. either at grain contacts between quartz grains (GTG coating) or at the interface between detrital quartz grains and the intergranular volume (GTI coating).Illitic GTI coatings affect syntaxial quartz overgrowth precipitation, as precipitation sites are locally blocked. The negative correlation between the grain coating coverage and quartz cement volumes support these findings across multiple sample sets, and they may locally preserve intergranular porosity. Illitic GTG coatings on the other hand enhance chemical compaction (i.e. pressure dissolution) and will reduce the IGV. The negative correlation between these two properties again underlines the negative effect of this process on reservoir properties. Studied samples from a deep Buntsandstein well in the central URG show low reservoir quality due to either intense quartz cementation (0.7–31.7%) or a high degree of mechanical and chemical compaction (IGV: 2.3–38.0%). Higher illitic GTG and GTI coating coverages play a substantial role in controlling reservoir quality development, as demonstrated by comparing data from other fluvio-eolian lithologies (Triassic Buntsandstein and Permian Rotliegendes) to results of this study. In relation to their respective burial histories, higher illitic GTI coating coverages always correlate with smaller syntaxial quartz cement volumes. Similarly, higher illitic GTG coating coverages always correlate with lower IGV values in the three compared sample series.As both, the precipitation of quartz cements and compaction, are a function of the burial history, i.e. effective stresses and experienced temperatures, understanding the interaction of both these processes may enable the prediction of reservoir properties in undrilled areas.

Pore Pressure Prediction for High-Pressure Tight Sandstone in the Huizhou Sag, Pearl River Mouth Basin, China: A Machine Learning-Based Approach

A growing number of large data sets have created challenges for the oil and gas industry in predicting reservoir parameters and assessing well productivity through efficient and cost-effective techniques. The design of drilling plans for a high-pressure tight-sand reservoir requires accurate estimations of pore pressure (Pp) and reservoir parameters. The objective of this study is to predict and compare the Pp of Huizhou Sag, Pearl River Mouth Basin, China, using conventional techniques and machine learning (ML) algorithms. We investigated the characteristics of low-permeability reservoirs by observing well-logging data sets and cores and examining thin sections under a microscope. In the reservoir zone, the average hydrocarbon saturation is 55%, and the average effective porosity is 11%. The tight sandstone reservoirs consist of fine- to extremely fine-grained argillaceous feldspathic sandstone. The mean absolute error for reservoir property prediction is 1.3%, 2.2%, and 4.8%, respectively, for effective porosity, shale volume, and water saturation. Moreover, the ML algorithm was employed to cross-check the validity of the prediction of Pp. Combining conventional and ML techniques with the core data demonstrates a correlation coefficient (R2) of 0.9587, indicating that ML techniques are the most effective in testing well data. This study shows that ML can effectively predict Pp at subsequent depths in adjacent geologically similar locations. Compared to conventional methods, a substantial data set and ML algorithms improve the precision of Pp predictions.

Real-Time Prediction of Petrophysical Properties Using Machine Learning Based on Drilling Parameters.

The prediction of rock porosity and permeability is crucial for assessing reservoir productivity and economic feasibility. However, traditional methods for obtaining these properties are time-consuming and expensive, making them impractical for comprehensive reservoir evaluation. This study introduces a novel approach to efficiently predict rock porosity and permeability for reservoir assessment by leveraging real-time machine learning models. Utilizing readily available drilling parameters, this approach offers a cost-effective alternative to traditional time-consuming methods to predict formation petrophysical parameters in real-time. The data set used in this study was collected from two vertical wells located in the Middle East. It encompasses drilling parameters such as the rate of penetration (ROP), gallons per minute (GPM), revolutions per minute (RPM), strokes per minute (SPP), torque, and weight on bit (WOB), along with the corresponding measurements of porosity (ϕ) and permeability (k) obtained through core analysis. Three machine learning models, namely, decision trees (DTs), random forest (RFs), and support vector machines (SVMs), were employed and evaluated for their effectiveness in predicting porosity and permeability. The results demonstrate promising performance across the different data sets. All three models achieved correlation coefficients (R) higher than 0.91 in predicting porosity. The RF model exhibited accurate predictions of permeability, achieving R values surpassing 0.92 in the various data sets. While the DT model displayed slightly lower performance, with the R-value decreasing to 0.88 in the testing data set, the SVM model suffered from overfitting, with R values dropping to 0.83 in the testing data set. The novelty of this work lies in the successful application of machine learning models to the real-time prediction of reservoir properties, providing a practical and efficient solution for the oil and gas industry. By achieving correlation coefficients exceeding 0.91 and showcasing the models' efficacy in a dynamic testing data set, this study paves the way for improved decision-making processes and enhanced exploration and production activities. The innovative aspect lies in the utilization of drilling parameters for timely and cost-effective estimation, transforming conventional reservoir evaluation methods.

A New Workflow for Estimating Reservoir Properties With Gradient Boosting Model and Joint Inversion Using MWD Measurements

Triple-combo logs are important measurements for estimating geological, petrophysical, and geomechanical properties. Unfortunately, wireline and advanced logging-while-drilling (LWD) logs are typically dropped from the formation evaluation plan for unconventional wells due to economic constraints or borehole instability risks. Available measurements are typically measurement-while-drilling (MWD) gamma ray (GR) logs, along with surface measurements such as weight on bit (WOB), rate of penetration (ROP), torque, rotation per minute (RPM), and differential pressure. The development of a robust and rapid model for predicting reservoir properties using this limited data set would be of high value for geological evaluation. Estimating such properties is a challenging task due to the nonlinear relationship between the available log data and unknown reservoir properties. A novel workflow that combines two sequential models is presented. First is a machine-learning (ML) algorithm to predict triple-combo logs from drilling dynamics and GR logs. To train the ML algorithm, well logs obtained from multiple wells located in the Eagle Ford and Permian Basins are scrutinized to identify important features. This process includes depth shifting, outlier detection, and feature selection, which allows for strategic hyperparameter tuning. Several regression algorithms are investigated, and it is found that gradient boosting algorithms yield superior prediction performance. Unlike random forest methods, boosting algorithms train predictors sequentially, each trying to correct its predecessor. After triple-combo logs are predicted from MWD logs, a physics-based joint inversion model is applied to estimate various reservoir properties. The trained model is deployed on a blind test well, and the predicted logs show excellent agreement compared to the corresponding triple-combo measurements. The multimineral inversion using predicted triple-combo logs yields a geologic model that is validated with elemental capture spectroscopy (ECS) measurements. Additionally, reconstructed logs from the forward model closely match measured logs by minimizing the cost function. Therefore, real-time estimated geological, petrophysical, and geomechanical properties can reveal complex geologic information and be used to mitigate uncertainty related to drilling optimization, reservoir characterization, development planning, and reserve estimation. Using the MWD logs to predict triple-combo logs followed by a joint inversion is an innovative approach for geological evaluation with a limited data set. The developed workflow can successfully provide (1) geologic lithofacies identification and rock typing, (2) more confidence in real-time drilling operation, (3) reservoir properties prediction, (4) missing log imputations and pseudo-log generation with forward modeling, (5) guidance for future logging and perforation, (6) reference for seismic quantitative interpretation (QI) and well tie, and (7) potentially massive computation time saving from days to minutes.

3D rock physics template-based probabilistic estimation of tight sandstone reservoir properties

Quantitative prediction of reservoir properties (e.g., gas saturation, porosity, and shale content) of tight reservoirs is of great significance for resource evaluation and well placements. However, the complex pore structures, poor pore connectivity, and uneven fluid distribution of tight sandstone reservoirs make the correlation between reservoir parameters and elastic properties more complicated and thus pose a major challenge in seismic reservoir characterization. We have developed a partially connected double porosity model to calculate elastic properties by considering the pore structure and connectivity to analyze these factors’ influences on the elastic behaviors of tight sandstone reservoirs. The modeling results suggest that the bulk modulus is likely to be affected by the pore connectivity coefficient, while the shear modulus is sensitive to the volumetric fraction of stiff pores. By comparing the model predictions with the acoustic measurements of the dry and saturated quartz sandstone samples, the volumetric fraction of stiff pores and the pore connectivity coefficient can be determined. Based on the calibrated model, we have constructed a 3D rock physics template that accounts for the reservoir properties’ impacts on the P-wave impedance, S-wave impedance, and density. The template combined with Bayesian inverse theory is used to quantify gas saturation, porosity, clay content, and their corresponding uncertainties from elastic parameters. The application of well-log and seismic data demonstrates that our 3D rock physics template-based probabilistic inversion approach performs well in predicting the spatial distribution of high-quality tight sandstone reservoirs in southwestern China.

Hierarchical automated machine learning (AutoML) for advanced unconventional reservoir characterization

Recent advances in machine learning (ML) have transformed the landscape of energy exploration, including hydrocarbon, CO2 storage, and hydrogen. However, building competent ML models for reservoir characterization necessitates specific in-depth knowledge in order to fine-tune the models and achieve the best predictions, limiting the accessibility of machine learning in geosciences. To mitigate this issue, we implemented the recently emerged automated machine learning (AutoML) approach to perform an algorithm search for conducting an unconventional reservoir characterization with a more optimized and accessible workflow than traditional ML approaches. In this study, over 1000 wells from Alberta’s Athabasca Oil Sands were analyzed to predict various key reservoir properties such as lithofacies, porosity, volume of shale, and bitumen mass percentage. Our proposed workflow consists of two stages of AutoML predictions, including (1) the first stage focuses on predicting the volume of shale and porosity by using conventional well log data, and (2) the second stage combines the predicted outputs with well log data to predict the lithofacies and bitumen percentage. The findings show that out of the ten different models tested for predicting the porosity (78% in accuracy), the volume of shale (80.5%), bitumen percentage (67.3%), and lithofacies classification (98%), distributed random forest, and gradient boosting machine emerged as the best models. When compared to the manually fine-tuned conventional machine learning algorithms, the AutoML-based algorithms provide a notable improvement on reservoir property predictions, with higher weighted average f1-scores of up to 15–20% in the classification problem and 5–10% in the adjusted-R2 score for the regression problems in the blind test dataset, and it is achieved only after ~ 400 s of training and testing processes. In addition, from the feature ranking extraction technique, there is a good agreement with domain experts regarding the most significant input parameters in each prediction. Therefore, it is evidence that the AutoML workflow has proven powerful in performing advanced petrophysical analysis and reservoir characterization with minimal time and human intervention, allowing more accessibility to domain experts while maintaining the model’s explainability. Integration of AutoML and subject matter experts could advance artificial intelligence technology implementation in optimizing data-driven energy geosciences.

A New Robust Weak Supervision Deep Learning Approach for Reservoir Properties Prediction in Malaysian Basin Field

The conventional seismic inversion approach is practical for operational work, as it only uses simple linearized algorithms and assumptions, but may be less applicable when dealing with a complex geological setting, especially in the Malay basin fields, as it may introduce non-linear noises and non-unique solutions. In the Malay basin, we also frequently struggle with a scarcity of reliable well data when performing seismic inversions. This makes finding an accurate prior model for inversion challenging and contributes to high uncertainty in properties’ estimation. Implementation of deep learning for seismic inversion has become routine and has shown increasing capability in addressing nonlinearity in inverse problems. In this work, we develop a robust approach to deep learning-based seismic inversion to predict elastic properties from seismic data. The approach incorporates synthetic well and seismic data generation from a set of rock physics knowledge called the rock physics library, which plays a significant role in dataset input for network training, validation, and testing to improve elastic properties in this field. The deep learning network architecture comprising UNET and RESNET-18 with weak supervision networks has proven to be useful to enhance computational work efficiency and prediction accuracy while handling the non-linearity of the data and the non-uniqueness of the solutions. We successfully validated the proposed method on actual field data from a clastic fluvial-dominated field in the Malay basin. Upon comparative analysis with the conventional method, both inversion results are comparable and capable of identifying the reservoir occurrence and distribution. The conventional method exposed the presence of scattered amplitude noises and prominent seismic imprints masking the reservoir. Meanwhile, the proposed method showed more stable, clearer definition and fewer noise inversion results but with a faster turn-around time and a more efficient workflow. There are substantial improvements of up to 31% in correlation accuracy achieved upon implementing the proposed method for elastic properties prediction compared to the conventional. The result implies that the proposed method can provide a good elastic properties prediction framework while addressing data limitations and sparsity issues in typical deep learning-based inversions.

Fundamental error in tree-based machine learning model selection for reservoir characterisation

Over the past two decades, machine learning techniques have been extensively used in predicting reservoir properties. While this approach has significantly contributed to the industry, selecting an appropriate model is still challenging for most researchers. Relying solely on statistical metrics to select the best model for a particular problem may not always be the most effective approach. This study encourages researchers to incorporate data visualization in their analysis and model selection process.To evaluate the suitability of different models in predicting horizontal permeability in the Volve field, wireline logs were used to train Extra-Trees, Ridge, Bagging, and XGBoost models. The Random Forest feature selection technique was applied to select the relevant logs as inputs for the models. Based on statistical metrics, the Extra-Trees model achieved the highest test accuracy of 0.996, RMSE of 19.54 mD, and MAE of 3.18 mD, with XGBoost coming in second. However, when the results were visualised, it was discovered that the XGBoost model was more suitable for the problem being tackled. The XGBoost model was a better predictor within the sandstone interval, while the Extra-Trees model was more appropriate in non-sandstone intervals. Since this study aims to predict permeability in the reservoir interval, the XGBoost model is the most suitable. These contrasting results demonstrate the importance of incorporating data visualisation techniques as an evaluation metric. Given the heterogeneity of the subsurface, relying solely on statistical metrics may not be sufficient to determine which model is best suited for a particular problem.

Bayesian linearized rock-physics amplitude-variation-with-offset inversion for petrophysical and pore-geometry parameters in carbonate reservoirs

Rock-physics amplitude-variation-with-offset (AVO) inversion aims at directly predicting reservoir properties from prestack seismic data. However, carbonate reservoir rocks often develop complex pore structures, which limits the application of conventional inversion methods with fixed pore-geometry parameters. We derive a PP-wave reflectivity equation in terms of porosity, water saturation, and pore-aspect ratio and we develop a rock-physics AVO inversion method to jointly estimate the petrophysical and pore-geometry parameters in carbonates. The reflectivity equation is obtained by linearizing the rock-physics model based on the differential effective medium model and Gassmann’s equation using Taylor series approximation, and we combine the linearization with the Aki-Richards equation. Thanks to the linearized model, the analytical solution to the inverse problem is derived using Bayesian linear theory. Our solution combines simulated annealing to estimate the prior information of the pore geometry and the iterative Bayesian inversion to update the posterior model. The method is tested on a benchmark data set and compared with conventional methods with a fixed aspect ratio. The results show that our method improves the petrophysical results. The method is also compared with two-step and nonlinear inversion methods to demonstrate its advantages. The method is applied to a field seismic section from a carbonate reservoir with a complex pore structure to demonstrate its advantages in terms of prediction accuracy.

Lithology and Porosity Distribution of High-Porosity Sandstone Reservoir in North Adriatic Using Machine Learning Synthetic Well Catalogue

Reservoir characterization on offshore fields often faces specific challenges due to limited or unevenly distributed well data. The object of this study is the North Adriatic poorly consolidated clastic reservoir characterized by high porosity. The seismic data indicate notable differences in reservoir quality spatially. The only two wells on the field drilled the best reservoir area. Seismic data, seismic reservoir characterization, and accurate integration with scarce well data were crucial. This paper demonstrates how the application of machine learning algorithms, specifically a Deep Forward Neural Network (DFNN), and the incorporation of pseudo-well data into the reservoir characterization process can improve reservoir properties prediction. The methodology involves creating different reservoir porosity and thickness scenarios using pseudo-well data, synthetic pre-stack seismic data generation, seismic inversion, and DFNN utilization to improve porosity prediction. This study also highlights the importance of lithology discrimination in the geological model to better constrain reservoir properties distribution in the entire reservoir volume. Facies probability analysis was utilized to define interdependence between litho–fluid classes established from the well data and acoustic impedance volume. Apart from the field well data, seismic inversion results, and DFNN porosity volume as main inputs, acknowledgments from the neighboring fields also had an important role.

Advanced Elastic and Reservoir Properties Prediction through Generative Adversarial Network

The prediction of subsurface properties such as velocity, density, porosity, and water saturation has been the main focus of petroleum geosciences. Advanced methods such as Full Waveform Inversion (FWI), Joint Migration Inversion (JMI) and ML-Rock Physics are able to produce better predictions than their predecessors, but they still require tedious manual interpretation that is prone to human error. The research on these methods remains open as they suffer from technical limitations. As computing resources are becoming cheaper, the use of a single deep-generative adversarial network is feasible in predicting all these properties in a completely data-driven manner. In our proposed method of multiscale pix2pix applied to SEG SEAM salt data, we have managed to map from one input, which is seismic post-stack data, to several outputs of reservoir and elastic properties such as porosity, velocity, and density by using only one trained model and without having to manually interpret or pre-process the input data. With 90% accuracy of the results in the synthetic data testing, the method is worthy of being explored by the petroleum geoscience fraternity.

Quantitative characterization of shale gas reservoir properties based on BiLSTM with attention mechanism

Evaluating the potential of shale gas reservoirs is inseparable from reservoir properties prediction. Accurate characterization of total organic carbon, porosity and permeability is necessary to understand shale gas reservoirs. Seismic data can help to estimate these parameters in the area crossing-wells. We develop an improved deep learning method to achieve shale gas reservoir properties estimation. The relationship between elastic attributes and reservoir properties is built up by training a deep bidirectional long short-term memory network, which is suitable for time/depth sequence prediction, on the logging and core data. Except some commonly used technologies, such as layer normalization and dropout, we also introduce attention mechanism to further enhance the prediction accuracy. Besides, we propose to carry on the normal scores transform on the input features, which aims to make the relationship between inputs and targets clear and easy to learn. During the training process, we construct quantile loss function, then use Adam algorithm to optimize the network. Not only the characterization results, but also the confidence interval can be output that is meaningful for uncertainty analysis. The well experiment indicates that the method is promising for reducing prediction errors when training samples are insufficient. After analyzing in wells, the established model is acted upon seismic inverted elastic attributes to characterize shale gas reservoirs in the whole studied area. The estimation results coincide well with the actual development results, showing the feasibility of the novel method on the characterization for shale gas reservoirs.

Neural Network Modeling of Electromagnetic Prediction of Geothermal Reservoir Properties

Neural Network Modeling of Electromagnetic Prediction of Geothermal Reservoir Properties

Using the Shallow Strain Tensor to Characterize Deep Geologic Reservoirs

AbstractStoring and recovering water, carbon, and heat from geologic reservoirs is central to managing resources in a changing climate. We tested the hypothesis that the strain tensor caused by injecting or producing fluids can be measured at shallow depths and interpreted to advance understanding of underlying deep aquifers or reservoirs. Geodetic‐grade strainmeters were deployed at 30 m depth overlying the Bartlesville Formation, a 500‐m‐deep sandstone near Tulsa, OK. The strainmeters are 220 m east of injection well 9A completed in a permeable lens at the base of the Bartlesville Formation. Water was injected into well 9A at approximately 1.0 L/s during four tests that ranged in duration from a few hours to a few weeks. The horizontal strain increased (tension) and the circumferential strain was a few times larger than the radial strain. The vertical strain decreased (compression) during injection. Strain rates were approximately 100 nε/day during the first few hours, but the rates decreased and were approximately 10 nε/day during most of the tests. Four independent methods of poroelastic simulation and inversion predict reservoir properties and geometries that are similar to each other and consistent with independent information about the reservoir. All strain interpretations predict that a boundary to the permeable lens occurs beneath the vicinity of the strainmeters, which is consistent with core data from the site. The boundary of the permeable lens is located by matching the vertical, radial and circumferential strains, which demonstrates the value of measuring the strain tensor.

Quantitative Characterization of Shallow Marine Sediments in Tight Gas Fields of Middle Indus Basin: A Rational Approach of Multiple Rock Physics Diagnostic Models

For the successful discovery and development of tight sand gas reserves, it is necessary to locate sand with certain features. These features must largely include a significant accumulation of hydrocarbons, rock physics models, and mechanical properties. However, the effective representation of such reservoir properties using applicable parameters is challenging due to the complicated heterogeneous structural characteristics of hydrocarbon sand. Rock physics modeling of sandstone reservoirs from the Lower Goru Basin gas fields represents the link between reservoir parameters and seismic properties. Rock physics diagnostic models have been utilized to describe the reservoir sands of two wells inside this Middle Indus Basin, including contact cement, constant cement, and friable sand. The results showed that sorting the grain and coating cement on the grain’s surface both affected the cementation process. According to the models, the cementation levels in the reservoir sands of the two wells ranged from 2% to more than 6%. The rock physics models established in the study would improve the understanding of characteristics for the relatively high Vp/Vs unconsolidated reservoir sands under study. Integrating rock physics models would improve the prediction of reservoir properties from the elastic properties estimated from seismic data. The velocity–porosity and elastic moduli-porosity patterns for the reservoir zones of the two wells are distinct. To generate a rock physics template (RPT) for the Lower Goru sand from the Early Cretaceous period, an approach based on fluid replacement modeling has been chosen. The ratio of P-wave velocity to S-wave velocity (Vp/Vs) and the P-impedance template can detect cap shale, brine sand, and gas-saturated sand with varying water saturation and porosity from wells in the Rehmat and Miano gas fields, both of which have the same shallow marine depositional characteristics. Conventional neutron-density cross-plot analysis matches up quite well with this RPT’s expected detection of water and gas sands.

Reservoir Property Prediction in the North Sea Using Machine Learning

Reservoir Property Prediction in the North Sea Using Machine Learning

Neural Network Modeling of Electromagnetic Prediction of Geothermal Reservoir Properties

This work conducts neural network modeling of temperature, thermal conductivity, and permeability predictions for depths greater than those drilled, as well as for the immediate vicinity of the exploratory borehole. For this purpose, we use data from three boreholes drilled earlier in the Soultz-sous-Forêts geothermal site (France) and the results of the magnetotelluric sounding performed there. It is shown that the relative accuracy of the predictions depends significantly on the relationship between the depth of the drilled borehole and the target depth of the prediction. For instance, for all the examined parameters, prediction errors become less than 5% if the prediction is made for depths that do not exceed the borehole depth by more than two times. In this case, the average errors of temperature and thermal conductivity predictions for the vicinity of the drilled borehole were 3.6% and 6%, respectively. The obtained results justified a new scheme for predicting the thermophysical and porosity/permeability properties of rocks while drilling exploratory geothermal boreholes.

Acoustic-electrical properties and rock physics models for shale-oil formations: prediction of reservoir properties of interbedded sandstone and shale layers

Acoustic-electrical properties and rock physics models for shale-oil formations: prediction of reservoir properties of interbedded sandstone and shale layers

New direction for regional reservoir quality prediction using machine learning - Example from the Stø Formation, SW Barents Sea, Norway

Recently, the petroleum industry has focused on deeply buried reservoir discoveries and exploring potential CO2 storage sites close to existing infrastructure to increase the life span of already operating installations to save time and cost. It is therefore essential for the petroleum industry to find an innovative approach that exploits the existing core- and well log data to be successful in their endeavor of effectively characterizing and predicting reservoir quality. Continuous data sources (e.g. wireline logs) have a huge potential compared with expensive, time inefficient and sporadic data from cores in determining reservoir quality for use in a regional context. However, whereas core analysis offers in-depth knowledge about rock properties and diagenetic processes, continuous data sources can be difficult to interpret without a formation-specific framework. Here, we demonstrated how the pre-existing core data could be effectively used by integrating petrographic- and facies data with a pure predictive machine learning (ML) based porosity predictor. The inclusion of detailed core analysis is important for determining which reservoir parameter(s) that should be modeled and for the interpretation of model outputs. By applying this methodology, a framework for deducing lithological and diagenetic attributes can be established to aid reservoir quality delineation from wireline logs that can be used in frontier areas. With the ML porosity model, a Random Forest Regressor, the square of the correlation was 0.84 between predicted- and helium porosity test data over a large dataset consisting of 38 wells within the Stø Formation across the SW Barents Sea. By integrating the continuous ML porosity logs and core data, it was possible to differentiate three distinct bed types on wireline log responses within the Stø Formation. Particularly, the relationship between Gamma ray (GR) and porosity was effective in separating high porosity clean sand-, low porosity cemented clean sand and more clay and silt rich intervals. Additionally, in the P-wave velocity (VP) - density domain, separation of high porosity clean sand- and heavily cemented low porosity clean sand intervals were possible. The results also show that the ML derived porosity curves coincide with previously published and independent facies data from a selection of the wells included in the study. This demonstrates the applicability of the model in the region, because the Stø Formation has been described to exhibit similar lithological- and mineralogical properties over large parts of the Western Barents Sea area. Even though, continuous porosity data could be estimated from other sources like VP, neutron or density logs, this would generally require matrix and fluid information. This study demonstrated the effectiveness of the ML model in generating continuous porosity logs that are useful for characterizing and predicting reservoir properties in new wells. This methodology offers a workflow for exploiting already acquired core and well log data for frontier exploration that can be adapted to other formations and exploration scenarios worldwide.