The QQ Field in the Gulf of Suez is a mature, geologically complex with multiple stacked, faulted reservoirs, with commingled production between different reservoirs. This paper illustrates the power of an automated tool to perform systematic, rapid, and detailed assessment of the reservoir performance, identify the key recovery obstacles and prepare remedial plans to enable the reservoir to produce to its full potential. The well and reservoir data were processed to compute a series of metrics and key performance indicators at various levels (well, layer, reservoir, well groups, etc.). The tool has several automated modules to facilitate rapid, metric-driven reservoir assurance and management. These modules include: (i) well production/injection allocation, (ii) wells decline curve analysis including event-detection, (iii) pressure and voidage analysis, and (iv) Contact analysis. Using performance analytics, the study quickly identified ways to improve the health of the reservoir and maximize its value. The QQ Field predominantly produces from two formations: Nubia and Nezzazat. Furthermore, there are multiple sub-layers in each formation. Reliable flow unit allocation is critical to gauge contribution of each layer, identify the undrained areas of the reservoir, and locate future development opportunities. The flow unit allocation module incorporates all available data such as PLT/ILT data, completion history, permeability of each flow unit at well level, relative pressures, and water influx model. Based on the allocated production, the current recovery factors in Nubia and Nezzazat are approximately 60% and 20% respectively. Analysis of RFT data reveals good vertical communication across Nubia. However, in some areas there is clear pressure discontinuity across layers. The reservoir pressure has dropped below the bubble point in both formations. As a result, water injection was initiated. The pressure in all parts of Nubia was restored above bubble point. Aquifer influx is sufficient to support the current withdrawal rates and further water injection is unnecessary. However, Nezzazat has a significantly higher reservoir complexity and therefore, shows a large variation in pressure behavior. It needs water injection to maintain the reservoir pressure above the bubble point in all parts of the reservoir. Based on the flow-unit allocation, the voidage replacement ratio (VRR) was calculated for each area and each layer. Even though the overall VRR in the waterflooded areas is above one, the distribution of the injected water is uneven. Redistributing injected water and ensuring that all the areas and all the layers are adequately supported will help to maximize recovery. The prolonged dip in oil price demands extreme efficiency. Sound reservoir management must not require unreasonable time or manpower. The rapid, automated analysis enables quick identification of the key areas for improvement in reservoir management practices and maximize the value of the asset.

Depositional mechanisms of sediments and post-depositional process often cause spatial variation and heterogeneity in rock fabric, which can impact the directional dependency of petrophysical, electrical, and mechanical properties. Quantification of the directional dependency of the aforementioned properties is fundamental for the appropriate characterization of hydrocarbon-bearing reservoirs. Anisotropy quantification can be accomplished through numerical simulations of physical phenomena such as fluid flow, gas diffusion, and electric current conduction in porous media using multi-scale image data. Typically, the outcome of these simulations is a transport property (e.g., permeability). However, it is also possible to quantify the tortuosity of the media used as simulation domain, which is a fundamental descriptor of the microstructure of the rock. The objectives of this paper are (a) to quantify tortuosity anisotropy of porous media using multi-scale image data (i.e., whole-core CT-scan and micro-CT-scan image stacks) through simulation of electrical potential distribution, diffusion, and fluid flow, and (b) to compare electrical, diffusional, and hydraulic tortuosity. First, we pre-process the images (i.e., CT-scan images) to remove non-rock material visual elements (e.g., core barrel). Then, we perform image analysis to identify different phases in the raw images. Then, we proceed with the numerical simulations of electric potential distribution. The simulation results are utilized as inputs for a streamline algorithm and subsequent direction-dependent electrical tortuosity estimation. Next, we conduct numerical simulation of diffusion using a random walk algorithm. The distance covered by each walker in each cartesian direction is used to compute the direction-dependent diffusional tortuosity. Finally, we conduct fluid-flow simulations to obtain the velocity distribution and compute the direction-dependent hydraulic tortuosity. The simulations are conducted in the most continuous phase of the segmented whole-core CT-scan image stacks and in the segmented pore-space of the micro-CT-scan image stacks. Finally, the direction-dependent tortuosity values obtained with each technique are employed to assess the anisotropy of the evaluated samples. We tested the introduced workflow on dual energy whole-core CT-scan images and on smaller scale micro-CT-scan images. The whole-core CT-scan images were obtained from a siliciclastic depth interval, composed mainly by spiculites. Micro-CT-scan images we obtained from Berea Sandstone and Austin Chalk formations. We observed numerical differences in the estimates of direction-dependent electrical, diffusional, and hydraulic tortuosity for both types of image data employed. The highest numerical differences were observed when comparing electrical and hydraulic tortuosity with diffusional tortuosity. The observed differences were significant specially in anisotropic samples. The documented comparison provides useful insight in the selection process of techniques for estimation of tortuosity. The use of core-scale image data in the proposed workflow provides semi-continuous estimates of tortuosity and tortuosity anisotropy which is typically not attainable when using pore-scale images. Additionally, the semi-continuous nature of the tortuosity and tortuosity anisotropy estimates in whole-core CT-scan image data provides an excellent tool for the selection of core plugs coring locations.

This work shows an analysis of the total enrollment, new admissions, graduates and degrees awarded in Petroleum Engineering at Bachelor's level in the last 10 years in Mexico, according to the statistical yearbooks published by the Asociación Nacional de Universidades e Instituciones de Educación Superior (ANUIES), which will allow to know the expected behavior of the student enrollment for the next 5 years. The digital transformation is discussed to propose the strategy of education in the medium and long term. An analysis is carried out on the study plans of 11 universities that teach Petroleum Engineering and have an active student chapter of the Society of Petroleum Engineers (SPE) in Mexico, to know the subjects taught and detect possible modifications supported by the compression of automation elements such as data analysis and artificial intelligence. Actions taken by other universities in the world for transition from traditional education to an education aligned to new global needs are discussed as well. Results show that student enrollment exhibits a strong dependence on barrel prices, the higher the price per barrel, the greater the number of new students enrolled. Student ranges from 4,800 to 13,663 at their peak. Derived from the fall in oil prices in 2014, there is an average annual decrease of 12%. If the trend continues to be the same, there will be less than 3,000 Petroleum Engineering students in Mexico by 2025. There are 55 Petroleum Engineering schools in the country, Veracruz and Tabasco together account for 56.5% of graduates nationwide in 2019. Technological innovations and technical challenges should shape the curricula in the Petroleum Engineering degree. It was also detected that the current education contemplates the study of traditional technologies, so it is advisable to align efforts to improve and strengthen the curricula of the universities that already offer the degree, instead of opening new schools. Mexico's national fields face increasingly complex technical challenges, such as the production of mature fields, unconventional reservoirs, deep-water exploration, and heavy and extra-heavy oils. Current Petroleum Engineering students are the ones who in few years will make important decisions about the direction of the industry, so it is of great interest to know the quality and quantity of them, and focus efforts on addressing deficiencies in fields related to technical challenges.

At the early stages of field life, the subsurface project team operates under lack of information. Due to the high uncertainties, decisions at the exploration and appraisal stages are often influenced by cognitive distortion that leads to overestimation or underestimation of hydrocarbon reserves and, as a result, to suboptimal investment decisions. World practice allows us to identify the most common causes of cognitive bias: the team focus on the most provable according to their view scenario and may ignore data that contradicts the chosen scenario, the opinions of the team members differ in the choice of the most likely scenario, the team members work with geological and geophysical (G&G) data performing separate tasks and may miss important connections between various sources of information. The consequences of these cognitive distortions cause an increase in risk capital, the duration of exploration activities, and the choice of suboptimal field developmentstrategy resulting in a decrease in the effectiveness of the exploration program and the project as a whole. To reduce such risks, it is possible to attract subject matter experts with extensive experience to support the project team. But the amount of experts is limited and this approach cannot be implemented for the entire portfolio of exploration projects. As result of a research project of Gazpromneft in a partnership with IBM Research, an innovative approach was developed for the objective integration of geological and geophysical data. The main idea of this approach is to support the geologist's decisions by an intelligent assistant working on the principles of the modern theory of knowledge engineering. Using the generalized expert knowledge, the intelligent assistant impartially integrates disparate geological information into a set of conceptual geological models (scenarios, objectively evaluates their probabilities, and helps to plan optimal exploration/appraisal activities.

Reservoir characterization requires accurate prediction of multiple petrophysical properties such as bulk density (or acoustic impedance), porosity, and permeability. However, it remains a big challenge in heterogeneous reservoirs due to significant diagenetic impacts including dissolution, dolomitization, cementation, and fracturing. Most well logs lack the resolution to obtain rock properties in detail in a heterogenous formation. Therefore, it is pertinent to integrate core images into the prediction workflow. This study presents a new approach to solve the problem of obtaining the high-resolution multiple petrophysical properties, by combining machine learning (ML) algorithms and computer vision (CV) techniques. The methodology can be used to automate the process of core data analysis with a minimum number of plugs, thus reducing human effort and cost and improving accuracy. The workflow consists of conditioning and extracting features from core images, correlating well logs and core analysis with those features to build ML models, and applying the models on new cores for petrophysical properties predictions. The core images are preprocessed and analyzed using color models and texture recognition, to extract image characteristics and core textures. The image features are then aggregated into a profile in depth, resampled and aligned with well logs and core analysis. The ML regression models, including classification and regression trees (CART) and deep neural network (DNN), are trained and validated from the filtered training samples of relevant features and target petrophysical properties. The models are then tested on a blind test dataset to evaluate the prediction performance, to predict target petrophysical properties of grain density, porosity and permeability. The profile of histograms of each target property are computed to analyze the data distribution. The feature vectors are extracted from CV analysis of core images and gamma ray logs. The importance of each feature is generated by CART model to individual target, which may be used to reduce model complexity of future model building. The model performances are evaluated and compared on each target. We achieved reasonably good correlation and accuracy on the models, for example, porosity R2=49.7% and RMSE=2.4 p.u., and logarithmic permeability R2=57.8% and RMSE=0.53. The field case demonstrates that inclusion of core image attributes can improve petrophysical regression in heterogenous reservoirs. It can be extended to a multi-well setting to generate vertical distribution of petrophysical properties which can be integrated into reservoir modeling and characterization. Machine leaning algorithms can help automate the workflow and be flexible to be adjusted to take various inputs for prediction.

This study investigated the ability to produce accurate multiphase flow profiles simulating the response of producing reservoirs, using Generative Deep Learning (GDL) methods. Historical production data from numerical simulators were used to train a GDL model that was then used to predict the output of new wells in unseen locations. This work describes a procedure in which data analysis techniques are used to gain insight into reservoir flow behavior at a field level based on existing historical data. The procedure includes clustering, dimensionality reduction, correlation, in addition to novel interpretation methodologies that synthesize the results from reservoir simulation output, characterizing flow conditions. The insight was then used to build and train a GDL algorithm that reproduces the multiphase reservoir behavior for unseen operational conditions with high accuracy. The trained algorithm can be used to further generate new predictions of the reservoir response under operational conditions for which we do not have previous examples in the training data set. We found that the GDL algorithm can be used as a robust multiphase flow simulator. In addition, we showed that the physics of flow can be captured and manipulated in the GDL latent space after training to reproduce different physical effects that did not exist in the original training data set. Applying the methodology to the problem of determining multiphase production rate from new producing wells in undrilled locations showed positive results. The methodology was tested successfully in predicting multiphase production under different scenarios including multiwell channelized and heterogeneous reservoirs. Comparison with other shallow supervised algorithms demonstrated improvements realized by the proposed methodology, compared to existing methods. The study developed a novel methodology to interpret both data and GDL algorithms, geared towards improving reservoir management. The method was able to predict the performance of new wells in previously undrilled locations without using a reservoir simulator.

Well testing is one of the vital process as part of reservoir performance monitoring. As field matures with increase in number of well stock, testing becomes tedious job in terms of resources (MPFM and test separators) and this affect the production quota delivery. In addition, the test data validation and approval follow a business process that needs up to 10 days before to accept or reject the well tests. The volume of well tests conducted were almost 10,000 and out of them around 10 To 15 % of tests were rejected statistically per year. The objective of the paper is to develop a methodology to reduce well test rejections and timely raising the flag for operator intervention to recommence the well test. This case study was applied in a mature field, which is producing for 40 years that has good volume of historical well test data is available. This paper discusses the development of a data driven Well test data analyzer and Optimizer supported by artificial intelligence (AI) for wells being tested using MPFM in two staged approach. The motivating idea is to ingest historical, real-time data, well model performance curve and prescribe the quality of the well test data to provide flag to operator on real time. The ML prediction results helps testing operations and can reduce the test acceptance turnaround timing drastically from 10 days to hours. In Second layer, an unsupervised model with historical data is helping to identify the parameters that affecting for rejection of the well test example duration of testing, choke size, GOR etc. The outcome from the modeling will be incorporated in updating the well test procedure and testing Philosophy. This approach is being under evaluation stage in one of the asset in ADNOC Onshore. The results are expected to be reducing the well test rejection by at least 5 % that further optimize the resources required and improve the back allocation process. Furthermore, real time flagging of the test Quality will help in reduction of validation cycle from 10 days hours to improve the well testing cycle process. This methodology improves integrated reservoir management compliance of well testing requirements in asset where resources are limited. This methodology is envisioned to be integrated with full field digital oil field Implementation. This is a novel approach to apply machine learning and artificial intelligence application to well testing. It maximizes the utilization of real-time data for creating advisory system that improve test data quality monitoring and timely decision-making to reduce the well test rejection.

Well integrity has become a crucial field with increased focus and being published intensively in industry researches. It is important to maintain the integrity of the individual well to ensure that wells operate as expected for their designated life (or higher) with all risks kept as low as reasonably practicable, or as specified. Machine learning (ML) and artificial intelligence (AI) models are used intensively in oil and gas industry nowadays. ML concept is based on powerful algorithms and robust database. Developing an efficient classification model for well integrity (WI) anomalies is now feasible because of having enormous number of well failures and well barrier integrity tests, and analyses in the database. Circa 9000 dataset points were collected from WI tests performed for 800 wells in Gulf of Suez, Egypt for almost 10 years. Moreover, those data have been quality-controlled and quality-assured by experienced engineers. The data contain different forms of WI failures. The contributing parameter set includes a total of 23 barrier elements. Data were structured and fed into 11 different ML algorithms to build an automated systematic tool for calculating imposed risk category of any well. Comparison analysis for the deployed models was performed to infer the best predictive model that can be relied on. 11 models include both supervised and ensemble learning algorithms such as random forest, support vector machine (SVM), decision tree and scalable boosting techniques. Out of 11 models, the results showed that extreme gradient boosting (XGB), categorical boosting (CatBoost), and decision tree are the most reliable algorithms. Moreover, novel evaluation metrics for confusion matrix of each model have been introduced to overcome the problem of existing metrics which don't consider domain knowledge during model evaluation. The innovated model will help to utilize company resources efficiently and dedicate personnel efforts to wells with the high-risk. As a result, progressive improvements on business, safety, environment, and performance of the business. This paper would be a milestone in the design and creation of the Well Integrity Database Management Program through the combination of integrity and ML.

New water treatment facilities in the Gulf of Mexico include a seawater Sulfate Removal Unit (SRU) to mitigate reservoir souring and scaling. The general industry sulfate target for offshore SRU is usually 20 mg/L or even 40 mg/L; however, some facilities may require <10 mg/L of sulfate in injection water, which makes water quality monitoring more critical and challenging. Current industrial practice relies on only pressure drop and a constant cleaning interval frequency to perform SRU maintenance which may result in reduced membrane life due to frequency cleaning or severe membrane fouling without the capability to predict fouling based on process conditions. The machine learning techniques applied will fill the gap and deliver a prediction model based on both simulation and real-time field data. This model will track and monitor the system key performance indicators (KPIs) including pressure, membrane fouling factor (FF), permeate sulfate concentration etc. The monitoring and prediction of these KPIs provide estimates on when the next maintenance procedure is required, track membrane system status for troubleshooting and actions, and optimize membrane performance by tuning operation conditions.

In gas wells, decreased/unstable production can occur due to difficult-to-predict dynamic effects resulted from late-life phenomena, such as liquid loading and flooding. To minimize the negative impact of these effects, maximize production and extend the wells’ lifetime, wells are often operated in an intermittent production regime. The goal of this work is to find the optimum production and shut-in cycles to maximize intermittent gas production as a decision support to operators. A framework suitable for single and multiple wells was developed by coupling a Deep Learning forward model trained on historical data with a population-based global optimizer, Particle Swarm Optimization (PSO). The forward model predicts the production rates and wellhead pressure during production and shut-in conditions, respectively. The PSO algorithm optimizes the operational criteria given operational and environmental objectives, such as maximizing production, minimizing start-up/shut-in actions, penalizing emissions under several constraints such as planned maintenances and meeting a contract production value. The accuracy of the Deep Learning models was tested on synthetic and field data. On synthetic data, mature wells were tested under different reservoir conditions such as initial water saturation, permeability and flow regimes. The relative errors in the predicted total cumulative production ranged between 0.5 and 4.6% for synthetic data and 0.9% for field data. The mean errors for pressure prediction were of 2-3 bar. The optimization framework was benchmarked for production optimization and contract value matching for a single-well (on field data) and a cluster of wells (synthetic data). Single-well production optimization of a North Sea well achieved a 3% production increase, including planned maintenances. Production optimization for six wells resulted in a 21% production increase for a horizon of 30 days, while contract value matching yielded 29/30 values within 3% of the target. The most optimum, repeatable and computationally efficient results were obtained using critical pressure/gas flowrates as operational criteria. This could enable real-time gas production optimization and operational decision-making in a wide range of well conditions and operational requirements.

One of the responsibilities of a directional driller (DD) is the computation of the current bit position given the last survey station measurement, and with that information calculate the path back to plan if directional correction is needed. Having only a few minutes during a drilling connection to perform these calculations, the DD is limited to compute only a handful of possible paths that will be presented to the Drilling Engineer/Company Man. With this information, the Company Man will decide which path to follow. The present work aims to develop a computer algorithm that replicates the field knowledge of DDs but can compute hundreds of paths in less than one minute. In addition, since the objective of the trajectory correction may differ, the algorithm also can optimize for one of three goals: maximum rate of penetration (ROP), minimum tortuosity in the path, or maximum footage in the drilling target window. The paper presents examples of four different path recommendations in the lateral portion of a horizontal well. The results show the optimum recommended paths for the same position for a specific optimization goal. Finally, a comparison between the running time and number of paths computed is presented. All results were obtained during the validation tests of the algorithm.

Objectives/Scope (25 - 50) PDO is in the process of transforming its well and urban planning by adopting digital technologies and Artificial Intelligence (AI) to improve organizational efficiency and maximize business value through faster quality decision. In 2020, PDO collaborated with a third-party contractor to provide a novel solution to an industry-wide problem: "how to effectively plan 100's of wells in a congested brownfield setting? ". Business Transformation This paper describes an innovative AI-assisted well planning method that is a game-changer for well planning in mature fields, providing efficiency in urban and well trajectory planning. It was applied in one of PDO's most congested fields with a targeted infill of 43m well spacing. The novel well planning method automatically designs and optimizes well trajectories for 100-200 new wells while considering surface, subsurface and well design constraints. Existing manual workflows in the industry are extremely time consuming and sequential (multiple man-months of work) - particularly for fields with a congested subsurface (350+ existing wells in this case) and surface (limited options for new well pads). These conventional and sequential ways of working are therefore likely to leave value on the table because it is difficult to find 100+ feasible well trajectories, and optimize the development in an efficient manner. The implemented workflow has the potential to enable step change in improvements in time and value for brownfield well and urban planning for all future PDO developments. Innovation The innovative AI assisted workflow, an industry first for an infill development of this size, evaluates, generates and optimizes from thousands of drillable trajectories to an optimized set for the field development plan (based on ranked value drivers, in this case, competitive value, cost and UR). The workflow provides a range of drillable trajectories with multi-scenario targets and surface locations, allowing ranking, selection and optimization to be driven by selected metrics (well length, landing point and/or surface locations). The approach leads to a step change reduction in cycle time for well and urban planning in a complex brownfield with 100-200 infill targets, from many months to just a few weeks. It provides potential game-changing digital solutions to the industry, enabling improved performance, much shorter cycle times and robust, unbiased well plans. The real footprint and innovation from this AI-assisted workflow is the use of state-of-the-art AI to enhance team collaboration and integration, supporting much faster and higher quality field development decisions. Application of Technology This paper describes a novel solution to integrated well planning. This is a tangible example of real digital transformation of a complex, integrated and multi-disciplinary problem (geologists, well engineers, geomatics, concept engineers and reservoir engineers), and only one of very few applied use cases in the industry. This application also gives an example of "augmented intelligence", i.e. how AI can be used to truly support integrated project teams, while the teams remain fully in control of the ultimate decisions. The success of this approach leans on the integrated teamwork across multiple technical disciplines, not only involving PDO's resources, but also WhiteSpace Energy as a 3 rd party service provider. The enhanced collaboration allowed all parties to highlight their constraints in an integrated way from the start, strengthening the technical discussion between disciplines and learning from each constraint impact and dependencies. (e.g. dog leg severity). In summary, the change in process flow moving from a sequential well planning and urban planning method to an iterative and fast AI solution – including all technical considerations from beginning represented for PDO an added value of over 6 months of direct cycle time HC acceleration.

In unconventional reservoirs, optimal completion controls are essential to improving well productivity and reducing costs. In this article, we propose a statistical model to investigate associations between shale oil production and completion parameters (e.g., completion lateral length, total proppant, number of hydraulic fracturing stages), while accounting for the influence of spatially heterogeneous geological conditions on hydrocarbon production. We develop a non-parametric regression method that combines a generalized additive model with a fused LASSO regularization for geological homogeneity pursuit. We present an alternating augmented Lagrangian method for model parameter estimations. The novelty and advantages of our method over the published ones are a) it can control or remove the heterogeneous non-completion effects; 2) it can account for and analyze the interactions among the completion parameters. We apply our method to the analysis of a real case from a Permian Basin US onshore field and show how our model can account for the interaction between the completion parameters. Our results provide key findings on how completion parameters affect oil production in that can lead to optimal well completion designs.

The energy industry has been transformed considerably in the last years. Sustainable development of oil and gas reservoir has become a major driver for these energy companies, and strengthened the focus to maximize hydrocarbon extraction while minimizing the associated carbon footprint. The focus has been further on maximizing efficiency and waste reduction in order to enhance profitability of projects. Challenges still remain in terms of that the carbon emissions from oilfield operations, related to the production, disposal and utilization of water and hydrocarbons, may be significant and the objective of increasing production has to be traded off in many instances against the quest for reducing carbon emissions. The fourth industrial revolution has brought new opportunities for companies to enhance decision making in their upstream development and optimize their recovery potential while minimizing the carbon footprint and associated cost. In this work, we present a smart approach for optimizing recovery while minimizing the carbon footprint of a reservoir in terms of the associated development and production activities. We use an advanced nonlinear autoregressive neural network approach integrated with time-lapse electromagnetic monitoring data to forecast production and carbon emissions from the reservoir in real-time, under uncertainty. The artificial intelligence approach also allows to investigate a circular carbon approach, where the produced greenhouse gases are re-injected into the well, while at the same time adjusting water injection levels. This allows to forecast and analyze the impact of a circular development plan. We tested the AI framework on a synthetic reservoir encompassing a complex carbonate fracture system and well setup. The carbon emissions were forecasted in real-time based on the previous production rates and the defined injection levels. The forecasted carbon emissions were then integrated into an optimization technique, in order to adjust injection levels to minimize water cut and overall carbon emissions, while optimizing production rates. Results were promising and highlighted the potential significant reductions in carbon emissions for the studied synthetic reservoir case. Moreover, the deployment of deep electromagnetic surveys was proved particularly beneficial as a deep formation evaluation monitoring method for tracking the injected waterfront inside the reservoir and optimizing the sweep efficiency, while minimizing the inefficient use of water injection. Accordingly, such integrated AI approach has a twofold benefit: maximizing the hydrocarbon productivity, while minimizing the water consumption and associated carbon emissions. Such framework represents a paradigm shift in reservoir management and improved oil recovery operations under uncertainty. It proposes an innovative integrated methodology to reduce the carbon footprint and attain a real-time efficient circular development plan.

Technological advances unveil a dual reality in the oil and gas Industry. On one hand, the benefits of blockchain and artificial intelligence (AI), among others, has arrived to revolutionize the industry. On the other hand, industry professionals remain trapped in bureaucratic processes that undermine their performance. The diagnosis: knowledge workers, responsible for optimizing the recovery and economic performance of the fields, are the missing link in the digital transformation chain. They are suffering the digitalization of the status quo. This paper puts forward a broad digital transformation framework designed to increase the knowledge worker's productivity. Digital transformation is not just about the implementation and use of cutting-edge technologies. It is also the response to digital trends, and about adopting new processes and redesigning existing ones to compete effectively in an increasingly digital world. Prioritizing technology as the ultimate goal puts the business processes and the knowledge workers aside from the discussion. The key to this proposal is rethinking the business model according to the possibilities of new technologies based on a six-dimension scheme: Corporate strategy: It defines the long-term vision and investment criteria for value creation. Technology is an element within a business scheme that should not be analyzed in isolation. Digital strategy: Within the corporate strategy, what operational and strategic role does technology play? Should it only support the company's operation, or should it drive strategic reinvention? Culture: While digital transformation is the company's response to digital trends, culture is the muscle that provides (or not) the attributes required to succeed in this transformation endeavor. Innovation and creativity should be promoted as part of the company's DNA. Knowledge processes: A business model, built on new technologies, will necessarily impose new and automated practices. While the automation of physical processes is a fact, the automation of knowledge processes is the weakest link. Data governance: It defines the necessary conditions that guarantee the quality of the information and its strategic acquisition. Two elements are a must: the automation of processes, thereby avoiding arbitrariness in data management; and centralized databases, thereby eliminating data duplicity and criteria discrepancy. Data Science: At this point in the model, the company has efficient, automatic, and fast processes, assuring the quality and availability of the data from its conception to the final storage. Then, data scientists will have all the means, and a clear and aligned vision (corporate strategy) to extract meaningful insights for the business.

The Tuscaloosa Marine Shale (TMS) formation is a clay- and liquid-rich emerging shale play across central Louisiana and southwest Mississippi with recoverable resources of 1.5 billion barrels of oil and 4.6 trillion cubic feet of gas. The formation poses numerous challenges due to its high average clay content (50 wt%) and rapidly changing mineralogy, making the selection of fracturing candidates a difficult task. While brittleness plays an important role in screening potential intervals for hydraulic fracturing, typical brittleness estimation methods require the use of geomechanical and mineralogical properties from costly laboratory tests. Machine Learning (ML) can be employed to generate synthetic brittleness logs and therefore, may serve as an inexpensive and fast alternative to the current techniques. In this paper, we propose the use of machine learning to predict the brittleness index of Tuscaloosa Marine Shale from conventional well logs. We trained ML models on a dataset containing conventional and brittleness index logs from 8 wells. The latter were estimated either from geomechanical logs or log-derived mineralogy. Moreover, to ensure mechanical data reliability, dynamic-to-static conversion ratios were applied to Young's modulus and Poisson's ratio. The predictor features included neutron porosity, density and compressional slowness logs to account for the petrophysical and mineralogical character of TMS. The brittleness index was predicted using algorithms such as Linear, Ridge and Lasso Regression, K-Nearest Neighbors, Support Vector Machine (SVM), Decision Tree, Random Forest, AdaBoost and Gradient Boosting. Models were shortlisted based on the Root Mean Square Error (RMSE) value and fine-tuned using the Grid Search method with a specific set of hyperparameters for each model. Overall, Gradient Boosting and Random Forest outperformed other algorithms and showed an average error reduction of 5 %, a normalized RMSE of 0.06 and a R-squared value of 0.89. The Gradient Boosting was chosen to evaluate the test set and successfully predicted the brittleness index with a normalized RMSE of 0.07 and R-squared value of 0.83. This paper presents the practical use of machine learning to evaluate brittleness in a cost and time effective manner and can further provide valuable insights into the optimization of completion in TMS. The proposed ML model can be used as a tool for initial screening of fracturing candidates and selection of fracturing intervals in other clay-rich and heterogeneous shale formations.

Uniaxial Compressive Strength (UCS) and Tensile Strength (TS) are among the essential rock parameters required and determined for rock mechanical studies in Petroleum Engineering. However, the determination of such parameters requires some laboratory experiments, which may be time-consuming and costly at the same time. In order to estimate these parameters efficiently and in a short period, some mathematical tools have been used by different researchers. When regression tools proved to give good results only in the limited range of data used, machine learning methods proved to be very accurate in generating models that can cover a wide range of data. In this study, two machine learning models were used to predict the UCS and TS, Support Vector Regression optimized by Genetic Algorithm (GA-SVR) and Artificial Neural Networks (ANNs). The results were discussed for both uniaxial compressive strength and tensile strength in terms of coefficient of determination R 2 , root mean squared error (RMSE) and mean average error (MAE). First, for the case of UCS, values of 0.99 and 0.99, values of 3.41 and 2.9 and values of 2.43 and 1.9 were obtained for R 2 , RMSE and MAE for the ANN and GA-SVR, respectively. Second, for the TS, the same analogy was followed, a coefficient R 2 of 0.99 and 0.99, RMSE values of 0.41 and 0.45 and MAE values of 0.30 and 0.39 were obtained for ANNs and GA-SVR, respectively. The next step was to assess these models on a different dataset consisting of data obtained from Bakken Field in Williston Basin, North Dakota, United States. The models showed excellent results comparing to the correlations they were compared with, outperforming them in terms of R 2 , RMSE and MAE, giving the following results for ANN and SVR respectively, R 2 of 0.93, 0.92, RMSE of 9.54, 11.22 and MAE of 7.28, 9.24. The resultant conclusion of this work is that the use of machine learning algorithms can generate universal models which reduce the time and effort to estimate some complex parameters such as UCS and Tensile Strength.

Reservoir uncertainties, high water cut, completion integrity along with declining production are the major challenges of a mature field. These integrated with dying facilities and poor field production are key issues that each oil and gas company is facing these days. Arresting production decline is an inevitable objective, but with the existing techniques/steps involved, it becomes a cumbersome and exorbitant affair for the operators to meet their requirements. In addition, incompetent and flawed well data makes it more challenging to analyze mature fields. Although flow rate data is the most easily accessible data for mature fields, the absence of pressure data (flowing bottom-hole or wellhead pressure) remains a big obstacle for the application of conventional production enhancement and well screening strategies for most of the mature fields. A real-time optimization tool is thus constructed by developing a hybrid modelling technique that encapsulates Kriging and Fuzzy Logic to account for the imprecisions and uncertainties involved while identification of subsurface locations for production optimization of a mature field using only production data. The data from the existing wells in the field is used to generate a membership function based on its historical performance and productivity, thereby generating a spatial map of prospective areas, where secondary development operations can be taken up for production optimization.

The focus of this work is on developing a cognitive tool that predicts the most frequent HSE hazards with the highest potential severity levels. The tool identifies these risks using a natural language processing algorithm on HSE leading and lagging indicator reports submitted to an oilfield services company’s global HSE reporting system. The purpose of the tool is to prioritize proactive actions and provide focus to raise workforce awareness. A natural language processing algorithm was developed to identify priority HSE risks based on potential severity levels and frequency of occurrence. The algorithm uses vectorization, compression, and clustering methods to categorize the risks by potential severity and frequency using a formulated risk index methodology. In the pilot study, a user interface was developed to configure the frequency and the number of the prioritized HSE risks that are to be communicated from the tool to those employees who opted to receive the information in a given location. From this pilot study using data reported in the company’s online HSE reporting system, the algorithm successfully identified five priority HSE risks across different hazard categories based on the risk index. Using a high volume of reporting data, the risk index factored multiple coefficients such as severity levels, frequency and cluster tightness to prioritize the HSE risks. The observations at each stage of the developed algorithm are as follows: In the data cleaning stage, all stop words (such as a, and, the) were removed, followed by tokenization to divide text in the HSE reports into tokens and remove punctuation. In the vectorization stage, many vectors were formed using the Term Frequency - Inverse Document Frequency (TF-IDF) method. In the compression stage, an autoencoder removed the noise from the input data. In the agglomerative clustering stage, HSE reports with similar words were grouped into clusters and the number of clusters generated per category were in the range of three to five. The novelty of this approach is its ability to prioritize a location’s HSE risks using an algorithm containing natural language processing techniques. This cognitive tool treats reported HSE information as data to identify and flag priority HSE risks factoring in the frequency of similar reports and their associated severity levels. The proof of concept has demonstrated the potential ability of the tool. The next stage would be to test predictive capabilities for injury prevention.

Image-based characterization of rock fabric is critical for understanding recovery mechanisms in shale formations due to the significant multiscale nature of shale source rocks. Nanoscale imaging is particularly important for characterizing pore-scale structure of shales. Nanoimaging techniques, however, have a tradeoff between high-resolution/high-contrast sample-destructive imaging modalities and low-contrast/low-resolution sample-preserving modalities. Furthermore, acquisition of nanoscale images is often time-consuming, expensive, and requires signficant levels of expertise, resulting in small image datasets that do not allow for accurate quantification of petrophysical or morphological properties. In this work, we introduce methods for overcoming these challenges in image-based characterization of the fabric of shale source rocks using deep learning models. We present a multimodal/multiscale imaging and characterization workflow for enhancing non-destructive microscopy images of shale. We develop training methods for predicting 3D image volumes from 2D training data and simulate flow through the predicted shale volumes. We then present a novel method for synthesizing porous media images using generative flow models. We apply this method to several datasets, including grayscale and multimodal 3D image volume generation from 2D training images. Results from this work show that the proposed image reconstruction and generation approaches produce realistic pore-scale 3D volumes of shale source rocks even when only 2D image data is available. The models proposed here enable new capabilities for non-destructive imaging of source rocks and we hope will improve our ability to characterize pore-scale properties and phenomena in shales using image data.