The paper provides a technical overview of an operator's Real-Time Drilling (RTD) ecosystem currently developed and deployed to all US Onshore and Deepwater Gulf of Mexico rigs. It also shares best practices with the industry through the journey of building the RTD solution: first designing and building the initial analytics system, then addressing significant challenges the system faces (these challenges should be common in drilling industry, especially for operators), next enhancing the system from lessons learned, and lastly, finalizing a fully integrated and functional ecosystem to provide a one-stop solution to end users.
The RTD ecosystem consists of four subsystems as shown in architecture
RTD ecosystem architecture
RTD ecosystem architecture
All of these subsystems are fully integrated and interact with each other to function as one system, providing a one-stop solution for real-time drilling optimization and monitoring. This RTD ecosystem has become a powerful decision support tool for the drilling operations team. While it was a significant effort, the long term operational and engineering benefits to operators designing such a real-time drilling analytics ecosystem far outweighs the cost and provides a solid foundation to continue pushing the historical limitations of drilling workflow and operational efficiency during this period of rapid digital transformation in the industry.
Potapenko, Dmitriy (Schlumberger) | Theuveny, Bertrand (Schlumberger) | Williams, Ryan (Schlumberger) | Moncada, Katharine (Schlumberger) | Campos, Mario (Schlumberger) | Spesivtsev, Pavel (Schlumberger) | Willberg, Dean (Schlumberger)
Highly efficient multi-stage hydraulic fractured horizontal wellbores are the dominant completion method for many basins worldwide. One potential weakness of multi-stage hydraulic fracturing is that the later stages of the completion workflow – frac-plug drill out (FPDO) and flowback – cause large pressure fluctuations and transient flows through the perforation clusters that coincide with a period of low closure stress in the fractures. The proppant packs in the fractures during this period are fragile and prone to failure. Previously reported results show that flowback and initial production practices have a major impact on proppant production, maintenance and disposal costs and the subsequent well performance. In this paper the results from over 200 FPDO and flowback operations from the United States and Argentina are reviewed. These results show that maintaining a balanced flowrate during FPDO operations is critical for minimizing inadvertent damage to the hydraulic fracture network.
The FPDO flowrate balance is the difference between the coiled tubing injection and annular return flowrates. The magnitude and sign of the balance corresponds to the instantaneous flowrate through the open perforation clusters into or out of the hydraulic fracture network. A positive balance rate, or overbalance, injects fluid into the fracture system. A negative balance rate, or underbalance, produces stimulation or formation fluids from the fracture network. Sudden changes between these two regimes creates local flows that can be severe enough to flush large quantities of proppant out of the fractures. Our results show that high-frequency multiphase flowmeters simplify the process of maintaining balance (no inflow, no outflow). Furthermore, close monitoring of any imbalance that develops, and rapid control of the surface choke and injection rate, can provide for an efficient operation while protecting the integrity of the fracture system.
Early monitoring of flowback and production with a high frequency flowmeter was shown to be extremely useful technique for optimizing well productivity during well clean-up. This paper also shows how a dual energy gamma ray multiphase flowmeter successfully quantified proppant produced during FPDO and flowback. Examples of the dynamics of sand production are shown, as well as correlations to events of excessive underbalance conditions.
At the end of the paper we show that most of the highlighted problems can be solved through making changes to the well construction workflow and accounting for relationships between various well operations. Incorporation of this workflow enables early prediction of well performance issues and their efficient resolution.
Hydrocarbon production from shale formation has become an essential part of the global energy supply in the past decade. The life of a project in an unconventional play significantly depends on the prediction of Estimated Ultimate Recovery (EUR). However, the conventional methodology to predict EUR becomes less accurate for shale formations, which significantly affects the economics returns of projects in unconventional plays. The objective of this article is to investigate the most important independent variables, including petrophysics and completion parameters, to estimate EUR by the machine learning algorithm. A novel machine learning model based on Random Forest Regression is introduced to predict EUR and to rank the importance of the independent variables.
In this article, production/petrophysics/engineering/ data with more than 25 variables from 4000 wells in Eagle Ford is summarized for analysis. The data is collected from production monitoring, well logging, well testing, seismic interpretation and lab experiments. This paper has three major components. Firstly, a multivariate linear regression model is created to predict the overall EUR. Secondly, the spatial autocorrelation analysis is carried out to identify whether spatial variables could affect the accuracy of the multivariate regression model. Thirdly, the Random Forest Regression models are trained to examine their reliability in predicting EUR with spatially autocorrelated data. The importance of key predictors is also identified. The final models are tuned with optimized hyperparameters. Through the article, the predictive capabilities of each Random Forest Regression model are discussed in detail to understand the physics behind unconventional hydrocarbon production mechanisms.
The results and workflow presented in this paper are insightful and novel. Firstly, we test the multivariate regression analysis with all the petrophysics and completion variables using the backward elimination method. This widely used model has a limitation of excluding the spatial information. In order to identify the impact of spatial variable, we calculate the Moran's Index and find out that the data in this study is clustered or spatially autocorrelated. The p-value for EUR, Oil EUR and Gas EUR are 0.000002, 0.000000 and 0.12, which all reject the null hypothesis that the data is randomly distributed. To include the spatial information in the prediction, we use advanced machine learning technology, Random Forest, to predict the EUR with a combination of petrophysics, completion variables and spatial information. The key variables to predict EUR, Oil EUR and Gas EUR by the Random Forest Regression are identified. However, the importance of the key variables to predict Oil EUR and Gas EUR are different. Therefore, we split the overall EUR Random Forest Regression model (57% explained) into two prediction models, one for Oil EUR prediction and one for Gas EUR prediction. The Gas EUR Random Forest Regression model has better performance (76% explained) compared to the Oil EUR Random Forest Regression model (60% explained).
This study provides a deeper understanding of unconventional hydrocarbon production prediction from a big data perspective, and proposes a novel and reliable machine-learning model to predict EUR to evaluate economic returns in Eagle Ford. Compared to the traditional multivariate regression model, our Random Forest Regression models are more reliable. In addition, the Random Forest technique is able to rank the importance of the relevant independent variables, and the rank of importance can be applied to guide and to improve data collection and model training for further study on this topic. The workflow presented in this article can be also used to train data for other unconventional resource plays.
The Powder River Basin has emerged over the past year as the latest source of oil production growth for the Lower 48. Companies ranging from a reborn Samson Resources to US onshore mainstays Devon, Chesapeake, and EOG are now betting on the basin to become a long-term core asset. Colorado’s industry lacks the size, variety, and Wild West characteristics of Texas, but that is precisely why the Centennial State’s oil production is surging to record levels. This paper describes a comprehensive field study of eight horizontal wells deployed in the stacked Niobrara and Codell reservoirs in the Wattenberg Field (Denver-Julesburg Basin).
The large independent put together a team of data scientists, software developers, and petrotechnical staff to create a forward-looking vision for how to use digital technology to solve problems. Baker Hughes is still a GE company, but it has partnered with a second company for artificial intelligence expertise, C3.ai. The deal is expected to speed the integration of AI into oilfield operations by the company which also markets GE’s device analytics platform, Predix. Marathon Oil says its shale fields are producing more oil and gas with less hands-on work from company personnel thanks to a growing arsenal of digital technologies and workflows. Malaysia’s Petronas, Shell Malaysia, and Thailand’s PTTEP are now in the midst of full-scale digital adoption.
The goal of our work was to maximize gas production and recovery from a horizontal appraisal well in the Mancos shale in New Mexico. This required a fracture design that would maximize perforation cluster efficiency and a lateral placement strategy that would maximize gas recovery. A key challenge was to design a fracture treatment that would overcome the extreme stress shadowing effects. Another key challenge was to optimize the lateral placement balancing multiple factors.
Fracture treatment simulations were completed for various designs. Fracture simulations indicated cluster efficiency could be dramatically improved by optimizing the way we pump the pad. A step-up technique for increasing pumping rates during the pad stage helped to initiate more fractures. Intra-stage diversion was utilized. Fracture simulations were performed to optimize the lateral placement. This required balancing multiple factors to access the highest gas-in-place (GIP) interval yet facilitate more fracture initiations per stage.
Fracture descriptions from the fracture simulations were input to a reservoir simulator to determine the optimal design. This paper will focus on the hydraulic fracture modeling.
This appraisal well was the most productive Mancos gas well ever delivered in the San Juan Basin. The 9,546’ lateral produced at a choke constrained plateau rate of about 13 MMscfd for 7 months and produced over 6 BCF in the first 20 months. A radioactive tracer log indicated an overall perforation cluster efficiency of 83%, a significant achievement in a shale with high stress shadowing.
The fracturing fluid design, diverter design and pumping techniques can be applied in many other shales as a low-cost way to increase perforation cluster efficiency, which will in turn result in higher production rates and higher cumulative recovery. Building on the success observed in the Mancos wells, BP and BPX Energy have subsequently utilized these techniques in other shale plays with success.
The concepts and workflow used to decide the optimal lateral placement is a well-defined approach that can be applied to other unconventional wells to increase hydrocarbon recovery.
Seismic attributes can be both powerful and challenging to incorporate into interpretation and analysis. Recent developments with machine learning have added new capabilities to multi-attribute seismic analysis. In 2018, Geophysical Insights conducted a proof of concept on 100 square miles of multi-client 3D data jointly owned by Geophysical Pursuit, Inc. (GPI) and Fairfield Geotechnologies (FFG) in the Denver-Julesburg Basin (DJ). The purpose of the study was to evaluate the effectiveness of a machine learning workflow to improve resolution within the reservoir intervals of the Niobrara and Codell formations, the primary targets for development in this portion of the basin.
The seismic data are from Phase 5 of the GPI/Fairfield Niobrara program in northern Colorado. A preliminary workflow which included synthetics, horizon picking and correlation of 28 wells was completed. The seismic volume was re-sampled from 2 ms to 1 ms. Detailed well time-depth charts were created for the Top Niobrara, Niobrara A, B and C benches, Fort Hays and Codell intervals. The interpretations, along with the seismic volume, were loaded into the Paradise® machine learning application, and two suites of attributes were generated, instantaneous and geometric. The first step in the machine learning workflow is Principal Component Analysis (PCA). PCA is a method of identifying attributes that have the greatest contribution to the data and that quantifies the relative contribution of each. PCA aids in the selection of which attributes are appropriate to use in a Self-Organizing Map (SOM). In this case, 15 instantaneous attribute volumes, plus the parent amplitude volume, were used in the PCA and eight were selected to use in SOMs. The SOM is a neural network-based machine learning process that is applied to multiple attribute volumes simultaneously. The SOM produces a non-linear classification of the data in a designated time or depth window.
For this study, a 60-ms interval that encompasses the Niobrara and Codell formations was evaluated using several SOM topologies. One of the main drilling targets, the B chalk, is approximately 30 feet thick; making horizontal well planning and execution a challenge for operators. An 8 X 8 SOM applied to 1 ms seismic data improves the stratigraphic resolution of the B bench. The neuron classification also images small but significant structural variations within the chalk bench. These variations correlate visually with the geometric curvature attributes. This improved resolution allows for precise well planning for horizontals within the bench. The 25 foot thick C bench and the 17 to 25 foot thick Codell are also seismically resolved via SOM analysis. Petrophysical analyses from wireline logs run in seven wells within the survey by Digital Formation; together with additional results from SOMs show the capability to differentiate a high TOC upper unit within the A marl which presents an additional exploration target. Utilizing 2d color maps and geobodies extracted from the SOMs combined with petrophysical results allows calculation of reserves for the individual reservoir units as well as the recently identified high TOC target within the A marl.
The results show that a multi-attribute machine learning workflow improves the seismic resolution within the Niobrara reservoirs of the DJ Basin and results can be utilized in both exploration and development.
Production from organic-rich shale petroleum systems is extremely challenging due to the complex rock and flow characteristics. An accurate characterization of shale reservoir rock properties would positively impact hydrocarbon exploration and production planning. We integrate large-scale geologic components with small-scale petrophysical rock properties to categorize distinct rock types in low porosity and low permeability shales. We then use this workflow to distinguish three rock types in the reservoir interval of the Niobrara shale in the Denver Basin of the United States: The Upper Chalks (A, B, and C Chalk), the Marls (A, B, and C Marl), and the Lower Chalks (D Chalk and Fort Hays Limestone). In our study area, we find that the Upper Chalk has better reservoir-rock quality, moderate source-rock potential, stiffer rocks, and a higher fraction of compliant micro- and nanopores. On the other hand, the Marls have moderate reservoir-rock quality, and a higher source rock potential. Both the Upper Chalks and the Marls should have major economic potentials. The Lower Chalk has higher porosity and a higher fraction of micro-and nanopores; however, it exhibits poor source rock potential. The measured core data indicates large mineralogy, organic-richness, and porosity heterogeneities throughout the Niobrara interval at all scale.
Unconventional petroleum systems are highly complex hydrocarbon resource plays both at the reservoir scale and at the pore scale (Aplin and Macquaker, 2011; Loucks et al., 2012; Hart et al., 2013; Hackley and Cardott, 2016). These organic-rich sedimentary plays, generally described as shale reservoirs, are composed of very fine silt-and clay-sized particles with grain sizes < 62.5 μm (Loucks et al., 2009; Nichols, 2009; Passey et al., 2010; Kuila et al., 2014; Saidian et al., 2014). They undergo extensive post-depositional diagenesis that transforms rock composition and texture, hydrocarbon storage and productivity, and reservoir flow features (Rushing et al., 2008; McCarthy et al., 2011; Jarvie, 2012; Milliken et al., 2012). Although some shale rock facies can retain depositional attributes during diagenesis, many critical reservoir properties, such as, mineralogy, pore structure, organic richness and present-day organic potential, etc., are significantly perturbed (Hackley and Cardott, 2016).
A fundamental component of a real-time drilling analytics system is automatic rig state detection. High frequency time series data (typically one data point per second) from multiple sensors on a drilling rig is processed and labeled with drilling states including: slide drilling, rotate drilling, pick up, in slips, and others. With labeled time-series data, the real-time system can derive operational KPIs (key performance indicators) with extremely high resolution, e.g., a statistical summary of rotary versus slide drilling time for the rig supervisor and drilling engineer to analyze efficiency. Later, such information can be leveraged to develop algorithms to detect abnormal drilling events and drive closed loop control.
A workflow was developed to clean and fill in any missing data. A rules-based model was then applied to classify the data into seventeen rig states. For the state “drilling”, a sub-classification was made to label rotate drilling and slide drilling. However, it is difficult to categorize “slide drilling” solely based on surface RPM due to top drive oscillation. In order to achieve acceptable accuracies, three machine learning models to classify “rotate drilling” and “slide drilling” were evaluated: Random Forest, Convolutional Neural Network (CNN), and a hybrid Convolutional Neural Network / Recurrent Neural Network (CNN/RNN).
Machine learning models were built for two basins, one model each, to accommodate different drilling styles. For the Delaware Basin, 10 wells with 9 million rows of data were chosen, and for the DJ Basin, 12 wells with 2 million rows of data were chosen. A legacy, rules-based algorithm was applied to label each row as rotate or slide drilling, and the misclassified records were manually corrected. The machine learning models were found to be far superior to rules-based models. For the wells in the training set, the accuracies of our rules-based models were 70% and 90% respectively, while the accuracies of our machine learning models were over 99%. The CNN model was proven to be the best model, excelling with high accuracy, short computation time, and scalability for big data applications.
The data cleaning, preprocessing, and machine learning algorithm has been deployed in Anadarko's Real-Time Drilling (RTD) ecosystem (Cao et al., 2018, 2019), which consists of four layers: a data source, analytics, data storage, and UI layer. KPIs, directional statistics, and engineering models are calculated in real-time and visualized through a web-based UI. This system can be accessed by any member of the drilling operations team. The system is regularly used to evaluate, compare, and optimize well performance. Future plans include pushing analytical models to the rig site with edge computing to facilitate drilling guidance and levels of automation. To our knowledge, this is the first time that a deep learning model has been used to analyze drilling time series data in a production real-time system.
Well-to-well interference is an increasingly discussed issue. Previously drilled and producing “parent” wells and recently drilled “child” wells are yielding a reduction in recovery rates in both short and long-term cases due to interference. A primary contributor to the variability in production is the presence of pressure sinks as the result of production depletion in the parent wells. Infill drilling will continue to occur in the development of unconventional plays, and it is crucial to gain an understanding of the impacts of well-to-well interference on hydraulic fracture generation.
This paper discusses a detailed approach to investigating well-to-well interference based on integrating hydraulic fracture modeling and reservoir simulation in two different formations, the Niobrara and Codell, in the Denver-Julesburg Basin. The geomechanical properties were calibrated by DFIT data and pressure matching of the parent well treatments. The resulting parent well fracture geometries were incorporated into a numerical reservoir model to determine the pressure depletion envelopes. The imported depletion model allows for the simulation of the child well treatments and associated impacts of the pressure sinks on fracture generation and the interaction between child and parent wells. The resulting depletion model provided a framework to investigate various methods to mitigate the effects of well-to-well communication in subsequent development. The developed workflow of well-to-well interference is applicable in understanding the effects of infill development in other producing basins.
The modeled child well treatments resulted in a clear indication of well-to-well communication with the parent wells that was attributable to pressure depletion. Actual field bottom-hole pressure measurements validated these results in the parent wells captured during the time of the child well treatments. Resulting proppant concentrations of the child well fractures indicated that the majority of the proppant transports towards the parent wells. Very little effective conductivity exists in the opposing direction of the depleted regions.
Slickwater treatment simulations indicate extremely asymmetric fractures that stay isolated to their respective target bench. For child wells in the same bench as the parent wells, fractures propagate directly toward the parent wells, with little to no fracture growth in the opposite direction.
Protection frac simulations indicate beneficial or detrimental results depending on the amount of repressurization that is achieved and the distance that the pressure transient extends into the reservoir. Re-pressurizing the reservoir surrounding the parent wells by 1,000 psi resulted in a reduction of well interference. A 500-psi scenario resulted in increased well interference between the parent wells. Several wells communicated with both parent wells due to the repressurization being insufficient to offset the depletion.
Natural repressurization of the reservoir to mitigate the effect of well interference was also investigated by using the reservoir model. Simulation of the parent wells being shut-in for three months prior to the child well treatments resulted in a pore pressure increase of only 280 psi. Based on the protection frac sensitivity of 500 psi, this is not a large enough repressurization to mitigate well-to-well interference successfully in the modeled scenarios.