|Theme||Visible||Selectable||Appearance||Zoom Range (now: 0)|
The conventional wisdom about artificial intelligence is that bigger is better. Consider neural networks, the software used for training algorithms that can discover patterns within information that humans may miss. Researchers are creating increasingly huge versions that can analyze massive amounts of data, and thus do things such as generate more realistic text in response to a query. But these enormous neural networks come with some costs, including making it difficult for researchers to figure out how and why the software makes its predictions and decisions. When these neural networks become so big, researchers can get lost attempting to make sense of the billions of interconnected calculations taking place.
The launch in Europe in mid-April of a new gas cloud imaging (GCI) system from US technology company Honeywell will reveal the ability of artificial intelligence (AI) systems to improve safety at oil and gas installations. The GCI system will provide automated and continuous monitoring for leaks of dangerous and polluting gases such as methane at oil and gas, chemical, and power-generation facilities across Europe. Part of the Honeywell Rebellion gas cloud imaging product portfolio, the Mini GCI is a compact device designed for congested areas and small sites. These systems can be placed throughout an industrial facility to continuously monitor for gas leaks and provide alerts as soon as they occur. The company said reducing gas emissions such as methane from hydrocarbon operations was one the most cost-effective and impactful methods to help reach global climate and environmental goals.
The proliferation of high-resolution datasets and decrease in sensor, storage, and computing costs are significantly extending our ability to grasp new concepts, improve predictions, and perform better field decisions. In a matter of a few years, we have also witnessed the consolidation of data science and machine learning as widespread disciplines that could help generate new technologies derived from data. The abundance of data and the persistence of elusive physical laws to satisfactorily explain the complexity of our assets and operations are promoting a renowned interest for finding ways to extend current model capabilities and decision workflow practices. Moreover, the current stringent economic environment and the increasing interest in pursuing cleaner, safer, and cheaper sources of energy are driving the need for more practical but more robust predictive and prescriptive models. Ultimately, the physics we know needs to rely on data to unmask the physics that we do not yet know.
Nicholson, A. Kirby (Pressure Diagnostics Ltd.) | Bachman, Robert C. (Pressure Diagnostics Ltd.) | Scherz, R. Yvonne (Endeavor Energy Resources) | Hawkes, Robert V. (Cordax Evaluation Technologies Inc.)
Abstract Pressure and stage volume are the least expensive and most readily available data for diagnostic analysis of hydraulic fracturing operations. Case history data from the Midland Basin is used to demonstrate how high-quality, time-synchronized pressure measurements at a treatment and an offsetting shut-in producing well can provide the necessary input to calculate fracture geometries at both wells and estimate perforation cluster efficiency at the treatment well. No special wellbore monitoring equipment is required. In summary, the methods outlined in this paper quantifies fracture geometries as compared to the more general observations of Daneshy (2020) and Haustveit et al. (2020). Pressures collected in Diagnostic Fracture Injection Tests (DFITs), select toe-stage full-scale fracture treatments, and offset observation wells are used to demonstrate a simple workflow. The pressure data combined with Volume to First Response (Vfr) at the observation well is used to create a geometry model of fracture length, width, and height estimates at the treatment well as illustrated in Figure 1. The producing fracture length of the observation well is also determined. Pressure Transient Analysis (PTA) techniques, a Perkins-Kern-Nordgren (PKN) fracture propagation model and offset well Fracture Driven Interaction (FDI) pressures are used to quantify hydraulic fracture dimensions. The PTA-derived Farfield Fracture Extension Pressure, FFEP, concept was introduced in Nicholson et al. (2019) and is summarized in Appendix B of this paper. FFEP replaces Instantaneous Shut-In Pressure, ISIP, for use in net pressure calculations. FFEP is determined and utilized in both DFITs and full-scale fracture inter-stage fall-off data. The use of the Primary Pressure Derivative (PPD) to accurately identify FFEP simplifies and speeds up the analysis, allowing for real time treatment decisions. This new technique is called Rapid-PTA. Additionally, the plotted shape and gradient of the observation-well pressure response can identify whether FDI's are hydraulic or poroelastic before a fracture stage is completed and may be used to change stage volume on the fly. Figure 1: Fracture Geometry Model with FDI Pressure Matching Case studies are presented showing the full workflow required to generate the fracture geometry model. The component inputs for the model are presented including a toe-stage DFIT, inter-stage pressure fall-off, and the FDI pressure build-up. We discuss how to optimize these hydraulic fractures in hindsight (look-back) and what might have been done in real time during the completion operations given this workflow and field-ready advanced data-handling capability. Hydraulic fracturing operations can be optimized in real time using new Rapid-PTA techniques for high quality pressure data collected on treating and observation wells. This process opens the door for more advanced geometry modeling and for rapid design changes to save costs and improve well productivity and ultimate recovery.
Abstract In this case study, we apply a novel fracture imaging and interpretation workflow to take a systematic look at hydraulic fractures captured during thorugh fracture coring at the Hydraulic Fracturing Test Site (HFTS) in Midland Basin. Digital fracture maps rendered using high resolution 3D laser scans are analyzed for fracture morphology and roughness. Analysis of hydraulic fracture faces show that the roughness varies systematically in clusters with average cluster separation of approximately 20' along the core. While isolated smooth hydraulic fractures are observed in the dataset, very rough fractures are found to be accompanied by proximal smoother fractures. Roughness distribution also helps understand the effect of stresses on fracture distribution. Locally, fracture roughness seems to vary with fracture orientations indicating possible inter-fracture stress effects. At the scale of stage lengths however, we see evidence of inter-stage stress effects. We also observe fracture morphology being strongly driven by rock properties and changes in lithology. Identified proppant distribution along the cored interval is also correlated with roughness variations and we observe strong positive correlation between proppant concentrations and fracture roughness at the local scale. Finally, based on the observed distribution of hydraulic fracture properties, we propose a conceptual spatio-temporal model of fracture propagation which can help explain the hydraulic fracture roughness distribution and ties in other observations as well.
Suarez-Rivera, Roberto (W. D. Von Gonten Laboratories) | Panse, Rohit (W. D. Von Gonten Laboratories) | Sovizi, Javad (Baker Hughes) | Dontsov, Egor (ResFrac Corporation) | LaReau, Heather (BP America Production Company, BPx Energy Inc.) | Suter, Kirke (BP America Production Company, BPx Energy Inc.) | Blose, Matthew (BP America Production Company, BPx Energy Inc.) | Hailu, Thomas (BP America Production Company, BPx Energy Inc.) | Koontz, Kyle (BP America Production Company, BPx Energy Inc.)
Abstract Predicting fracture behavior is important for well placement design and for optimizing multi-well development production. This requires the use of fracturing models that are calibrated to represent field measurements. However, because hydraulic fracture models include complex physics and uncertainties and have many variables defining these, the problem of calibrating modeling results with field responses is ill-posed. There are more model variables than can be changed than field observations to constrain these. It is always possible to find a calibrated model that reproduces the field data. However, the model is not unique and multiple matching solutions exist. The objective and scope of this work is to define a workflow for constraining these solutions and obtaining a more representative model for forecasting and optimization. We used field data from a multi-pad project in the Delaware play, with actual pump schedules, frac sequence, and time delays as used in the field, for all stages and all wells. We constructed a hydraulic fracturing model using high-confidence rock properties data and calibrated the model to field stimulation treatment data varying the two model variables with highest uncertainty: tectonic strain and average leak-off coefficient, while keeping all other model variables fixed. By reducing the number of adjusting model variables for calibration, we significantly lower the potential for over-fitting. Using an ultra-fast hydraulic fracturing simulator, we solved a global optimization problem to minimize the mismatch between the ISIPs and treatment pressures measured in the field and simulated by the model, for all the stages and all wells. This workflow helps us match the dominant ISIP trends in the field data and delivers higher confidence predictions in the regional stress. However, the uncertainty in the fracture geometry is still large. We also compared these results with traditional workflows that rely on selecting representative stages for calibration to field data. Results show that our workflow defines a better global optimum that best represents the behavior of all stages on all wells, and allows us to provide higher-confidence predictions of fracturing results for subsequent pads. We then used this higher confidence model to conduct sensitivity analysis for improving the well placement in subsequent pads and compared the results of the model predictions with the actual pad results.
ABSTRACT The industry is facing significant challenges due to the recent downturn in oil prices, particularly for the development of tight reservoirs. It is more critical than ever to 1) identify the sweet spots with less uncertainty and 2) optimize the completion-design parameters. The overall objective of this study is to quantify and compare the effects of reservoir quality and completion intensity on well productivity. We developed a supervised fuzzy clustering (SFC) algorithm to rank reservoir quality and completion intensity, and analyze their relative impacts on wells' productivity. We collected reservoir properties and completion-design parameters of 1,784 horizontal oil and gas wells completed in the Western Canadian Sedimentary Basin. Then, we used SFC to classify 1) reservoir quality represented by porosity, hydrocarbon saturation, net pay thickness and initial reservoir pressure; and 2) completion-design intensity represented by proppant concentration, number of stages and injected water volume per stage. Finally, we investigated the relative impacts of reservoir quality and completion intensity on wells' productivity in terms of first year cumulative barrel of oil equivalent (BOE). The results show that in low-quality reservoirs, wells' productivity follows reservoir quality. However, in high-quality reservoirs, the role of completion-design becomes significant, and the productivity can be deterred by inefficient completion design. The results suggest that in low-quality reservoirs, the productivity can be enhanced with less intense completion design, while in high-quality reservoirs, a more intense completion significantly enhances the productivity. Keywords Reservoir quality; completion intensity; supervised fuzzy clustering, approximate reasoning,tight reservoirs development
Abstract In previous frac designs, proppant tracer logs revealed poor proppant distribution between clusters. In this study, various technologies were utilized to improve cluster efficiency, primarily focusing on selecting perforations in like-rock, adjusting perforation designs and the use of diverters. Effectiveness of the changes were analyzed using proppant tracer. This study consisted of a group of four wells completed sequentially. Sections of each well were divided into completion design groups characterized by different perforating methodologies. Perforation placement was primarily driven by RockMSE (Mechanical Specific Energy), a calculation derived from drilling data that relates to a rock's compressive strength. Additionally, the RockMSE values were compared alongside three different datasets: gamma ray collected while drilling, a calculation of stresses from accelerometer data placed at the bit, and Pulsed Neutron Cross Dipole Sonic log data. The results of this study showed strong indications that fluid flow is greatly affected by rock strength as mapped with the RockMSE, with fluid preferentially entering areas with low RockMSE. It was found that placing clusters in similar rock types yielded an improved fluid distribution. Additional improved fluid distribution was observed by adjusting hole diameter, number of perforations and pump rate.
Wu, Yinghui (Silixa LLC) | Hull, Robert (Silixa LLC) | Tucker, Andrew (Apache Corp.) | Rice, Craig (Apache Corp.) | Richter, Peter (Silixa LLC) | Wygal, Ben (Silixa LLC) | Farhadiroushan, Mahmoud (Silixa Ltd.) | Trujillo, Kirk (Silixa LLC) | Woerpel, Craig (Silixa LLC)
Abstract Distributed fiber-optic sensing (DFOS) has been utilized in unconventional reservoirs for hydraulic fracture efficiency diagnostics for many years. Downhole fiber cables can be permanently installed external to the casing to monitor and measure the uniformity and efficiency of individual clusters and stages during the completion in the near-field wellbore environment. Ideally, a second fiber or multiple fibers can be deployed in offset well(s) to monitor and characterize fracture geometries recorded by fracture-driven interactions or frac-hits in the far-field. Fracture opening and closing, stress shadow creation and relaxation, along with stage isolation can be clearly identified. Most importantly, fracture propagation from the near to far-field can be better understood and correlated. With our current technology, we can deploy cost effective retrievable fibers to record these far-field data. Our objective here is to highlight key data that can be gathered with multiple fibers in a carefully planned well-spacing study and to evaluate and understand the correspondence between far-field and near-field Distributed Acoustic Sensing (DAS) data. In this paper, we present a case study of three adjacent horizontal wells equipped with fiber in the Permian basin. We can correlate the near-field fluid allocation across a stage down to the cluster level to far-field fracture driven interactions (FDIs) with their frac-hit strain intensity. With multiple fibers we can evaluate fracture geometry, the propagation of the hydraulic fractures, changes in the deformation related to completion designs, fracture complexity characterization and then integrate the results with other data to better understand the geomechanical processes between wells. Novel frac-hit corridor (FHC) is introduced to evaluate stage isolation, azimuth, and frac-hit intensity (FHI), which is measured in far-field. Frac design can be evaluated with the correlation from near-field allocation to far-field FHC and FHI. By analyzing multiple treatment and monitor wells, the correspondence can be further calibrated and examined. We observe the far-field FHC and FHI are directly related to the activities of near-field clusters and stages. A leaking plug may directly result in FHC overlapping, gaps and variations in FHI, which also can be correlated to cluster uniformity. A near-far field correspondence can be established to evaluate FHC and FHI behaviors. By utilizing various completion designs and related measurements (e.g. Distributed Temperature Sensing (DTS), gauges, microseismic etc.), optimization can be performed to change the frac design based on far-field and near-field DFOS data based on the Decision Tree Method (DTM). In summary, hydraulic fracture propagation can be better characterized, measured, and understood by deploying multiple fibers across a lease. The correspondence between the far-field measured FHC and FHI can be utilized for completion evaluation and diagnostics. As the observed strain is directly measured, completion engineering and geoscience teams can confidently optimize their understanding of the fracture designs in real-time.