Abstract Sucker rod pumping systems often require tubing anchors to secure the tubing string to the casing near the pump, minimizing damaging and tubing movement that reduces efficiency during operation. However, conventional tubing anchors can restrict the annular flow path, limiting production, reducing downhole gas separation efficiency, increasing flow slugging, and elevating fluid levels in the annulus. Increased slugging may also exacerbate solids-related issues, as slug flow is a primary mechanism for solids transport in horizontal wells. To address these challenges, a tubing anchor has been engineered for low operational risk and high flow, particularly for gassy sluggy horizontal wells. This tubing anchor maintains an unrestricted annular flow path and a full 2-7/8 in. External Upset End (EUE) tubing internal diameter, enabling the use of larger sucker rod pumps and placement above the pump. The anchor's innovative d esign allows setting and unsetting without tubing rotation, using simple up-or-down tubing string movements, therefore reducing operational risks associated with capillary chemical injection lines or placement in the wellbore curve. Oxy conducted a comprehensive, long-term evaluation of this tubing anchor, controlling variables and minimizing changes to the pump and bottomhole assembly (BHA). The program yielded statistically significant improvement in production across a meaningful sample set in the Permian Basin, demonstrating improved production performance and reliability. Key findings from the evaluation are presented, highlighting the observed mitigation of historical production decline trends following artificial lift conversion from ESPs to sucker rod pumping systems.
Abstract While many different KPIs have been developed over the years for evaluating well construction performance, the definition and utilization of a wellbore quality KPI have remained challenging due to its subjective nature, data wrangling issues, and human resource demands. This paper introduces an automated method for calculating a Wellbore Quality Index (WQI) using 1 Hz drilling data, ML-generated performance outputs and physics-based models, to assist decision-making for casing running de-risking and clean-out requirements. The WQI is calculated as a weighted summation of multiple features, where the weights are derived using Linear Discriminant Analysis (LDA). The WQI was validated for two applications: casing runnability prediction using 12 features derived from BHA runs prior to casing or liner installation, and overall wellbore quality post-analysis using 15 features including casing-related features. Most features focused on indicators related to pipe movement difficulty. The WQI was developed and evaluated using an initial population of 118 horizontal wells. The initial parameter list incorporated a study on Operator-defined Tortuosity Index (TI) to understand its possible impact on wellbore quality. NPT data associated with the WQI was considered during evaluation. Data quality checks identified missing or out-of-range values. For casing runnability prediction, the WQI achieved an in-sample Area Under the Curve (AUC) of 0.971 with a Leave-One-Out Cross-Validation (LOOCV) AUC of 0.906, demonstrating strong and generalizable discriminative performance across 104 hole sections. For overall wellbore quality, the index achieved a test AUC of 0.923 across 230 hole sections. In both cases, statistically significant separation between High and Low NPT groups was confirmed using the Mann-Whitney U test (p < 0.0001). The results support robust risk evaluation and inform decision-making regarding wellbore conditions in relation to pipe movement. Parameter selection is adaptable, enabling use across various basins, well types, and business units. Real-time data can be processed down to a bit run, hole section, and at the end of the well. This paper presents an automated method for determining wellbore quality using physics-based models and drilling performance calculations generated using machine learning models (Bayesian network). This approach offers clearer, more usable results for prediction, root cause analysis, risk assessment, and performance benchmarking. The index, along with the various subcomponents, provides clear guidance on actions to be taken prior to pulling the BHA out of the hole for a casing run.
Abstract Survey Accuracy plays a crucial role in optimizing well placement and minimizing uncertainty during drilling operations. Conventional MWD surveys are subject to multiple error sources that can lead to significant deviation from the actual wellbore position. The objective of this research is to demonstrate the importance and effectiveness of applying real-time high-fidelity (Hi-Fi) survey corrections to improve wellbore positioning accuracy and reduce the Ellipsoid of Uncertainty (EOU). Using Redundant or Relative Instrument Performance (RIP) testing as a validation method, the study compares corrected MWD surveys to high-resolution gyroscopic (HRG) data across a large set of wells. The data set consists of over 300 wells, with 201 of those wells utilizing advanced survey corrections in real-time. All 201 wells were reprocessed using a Hi-Fi trajectory solution, in which 15-foot survey points were injected between standard 95-foot survey intervals to improve resolution. Post-drilling, 15-foot-HRG surveys were available, providing a high-quality reference for validation. The RIP test statistically assesses the alignment between MWD and gyro surveys based on their respective error model. Additionally, deltas between the bottom-hole locations of the Hi-Fi and HRG surveys, with a small subset of MSA+SAG+IFR-only wells, were analyzed to quantify the impact of the corrections. The results demonstrate strong agreement between the Hi-Fi-corrected MWD surveys and the HRG references. Of 201 wells, 67% showed good agreement in the RIP test, 22% moderate agreement, and only 11% exhibited notable deviations, possibly due to improper setup or data-quality issues. More than 70% of wells were within 10 feet of the gyro reference. A two-dimensional EOU analysis indicated 95% containment, with half-widths of 20.6 ft horizontally and 13.8 ft vertically. When 26 outlier wells were excluded, the recalculated EOU values decreased to 13.5 ft and 10.6 ft, representing a 35% and 23% refinement in horizontal and vertical uncertainty, respectively. Also, Gaussian probability testing further confirmed that the residual positional differences followed a normal distribution with no directional bias. This study also highlights that TVD comparisons showed improvements in 58% of wells, ranging from 2 to over 6 ft. At the same time, horizontal displacements decreased by up to 40 ft when comparing Survey corrections to RAW MWD+IFR. Overall, Hi-Fi corrections achieved a 39% reduction in vertical uncertainty and a 16% reduction in horizontal uncertainty. This paper presents one of the most extensive validation studies comparing corrected MWD surveys to HRG data using RIP testing. By incorporating Hi-Fi survey corrections and post-drill HRG surveys across 201 wells, this study provides practicing engineers with practical, data-backed insight into the measurable benefits of real-time survey correction in the intermediate and curve sections.
Recent advances in generative networks have enabled new approaches to subsurface velocity model synthesis, offering a compelling alternative to traditional methods such as Full Waveform Inversion. However, these approaches predominantly rely on the availability of large-scale datasets of high-quality, geologically realistic subsurface velocity models, which are often difficult to obtain in practice. We introduce SAGE, a novel framework for statistically consistent proxy velocity generation from incomplete observations, specifically sparse well logs and migrated seismic images. During training, SAGE learns a proxy posterior over velocity models conditioned on both modalities (wells and seismic); at inference, it produces full-resolution velocity fields conditioned solely on migrated images, with well information implicitly encoded in the learned distribution. This enables the generation of geologically plausible and statistically accurate velocity realizations. We validate SAGE on both synthetic and field datasets, demonstrating its ability to capture complex subsurface variability under limited observational constraints. Furthermore, samples drawn from the learned proxy distribution can be leveraged to train downstream networks, supporting inversion workflows. Overall, SAGE provides a scalable and data-efficient pathway toward learning geological proxy posterior for seismic imaging and inversion. Repo link: https://github.com/slimgroup/SAGE.
Wireline formation pressure tests (pretests), are critical in reservoir characterization during exploration stage, providing reservoir pressure and mobility. We have demonstrated the application of a novel and fully automated digital pretest workflow in onshore, UAE. The use of intelligent real-time pretest classification combined with automated setting and retract of pretest probes led to significant reduction in overall test durations with increased consistency in results. These advancements allowed the reservoir engineers and asset managers to focus on interpretation of results, enhancing reservoir characterization and reducing operational risks This novel workflow is enabled by four automation stages combined and executed over a single command from a new customized acquisition system at wellsite: Automated setting of pretest probes, automated real time pretest classification (RTPC), automated pressure and mobility estimation and automated retract of probes. The new acquisition system is mounted with both logging tool electronic control algorithm and pretest interpretation algorithms, hence bringing the pretest interpretation to the wellsite minimizing the time loss due to delay in signal transmission from wellsite to the office. The first stage includes automated setting of probes in less than 30 seconds compared to conventional duration of 90 seconds. Next stage is automated pretest classification, where K-nearest neighbor (KNN) & Light grading boosting machine (LGBM) based AI/ML algorithms are used to classify the pretest type as valid test or invalid test. For valid tests, pressure stabilization is monitored over 10 second windows, and mobility is auto computed. For a tight test, the fourth automation stage is triggered, leading to retracting probes in 10 seconds with 95% accuracy. This process reduced manual interventions from 25 clicks to 5 per station, ensuring consistency in operations across multiple depths. The machine learning algorithm made the testing process more efficient and produced consistent, reliable pressure measurements for valid test and cut unnecessary time for tight tests. A total of 93 pretests were taken using a pre-planned automation algorithm executed over single command at each depth. The improvement in pretest efficiency and consistency was measured in time saving and data quality. Around 80% reduction in time spent on probe setting and retract was observed. The average pretest station time reduced from 13 minutes to less than 8 minutes, achieving a 38% overall time reduction. Tight tests and valid tests saw a 50%- and 40%-time reduction, leading to an overall time saving of over 7 hours during the pretest logging operation. Faster testing in openhole conditions also reduced wireline sticking risks. The optimized parameters enhanced data quality and consistency, resulting in the ability to obtain additional intervals with measurable pressure gradients. We showcased the application of AI & ML automation algorithms, tested and trained over 890 and 153 datasets respectively, and deployed on more than 20 real time pretest operations across the globe, before being used in this pretest operation. The real time pretest classification (RTPC) brought speed and consistency by reducing the coefficient of human error in such long pretest operations, leading to an efficient and standardized way of pressure testing.
