Artificial Intelligence Assisted Well Portfolio Optimization - An Automated Reservoir Management Advisory System to Maximize the Asset Value - Case Study from ADNOC Onshore

2021 ◽  
Author(s):  
Subba Ramarao Rachapudi Venkata ◽  
Nagaraju Reddicharla ◽  
Shamma Saeed Alshehhi ◽  
Indra Utama ◽  
Saber Mubarak Al Nuimi ◽  
...  
Keyword(s):  
Machine Learning ◽  
Hybrid Approach ◽  
Simulation Models ◽  
Production Optimization ◽  
Added Value ◽  
Risk Scores ◽  
Well Performance ◽  
Field Development ◽  
Advisory System ◽  
Gas Lift

Abstract Matured hydrocarbon fields are continuously deteriorating and selection of well interventions turn into critical task with an objective of achieving higher business value. Time consuming simulation models and classical decision-making approach making it difficult to rapidly identify the best underperforming, potential rig and rig-less candidates. Therefore, the objective of this paper is to demonstrate the automated solution with data driven machine learning (ML) & AI assisted workflows to prioritize the intervention opportunities that can deliver higher sustainable oil rate and profitability. The solution consists of establishing a customized database using inputs from various sources including production & completion data, flat files and simulation models. Automation of Data gathering along with technical and economical calculations were implemented to overcome the repetitive and less added value tasks. Second layer of solution includes configuration of tailor-made workflows to conduct the analysis of well performance, logs, output from simulation models (static reservoir model, well models) along with historical events. Further these workflows were combination of current best practices of an integrated assessment of subsurface opportunities through analytical computations along with machine learning driven techniques for ranking the well intervention opportunities with consideration of complexity in implementation. The automated process outcome is a comprehensive list of future well intervention candidates like well conversion to gas lift, water shutoff, stimulation and nitrogen kick-off opportunities. The opportunity ranking is completed with AI assisted supported scoring system that takes input from technical, financial and implementation risk scores. In addition, intuitive dashboards are built and tailored with the involvement of management and engineering departments to track the opportunity maturation process. The advisory system has been implemented and tested in a giant mature field with over 300 wells. The solution identified more techno-economical feasible opportunities within hours instead of weeks or months with reduced risk of failure resulting into an improved economic success rate. The first set of opportunities under implementation and expected a gain of 2.5MM$ with in first one year and expected to have reoccurring gains in subsequent years. The ranked opportunities are incorporated into the business plan, RMP plans and drilling & workover schedule in accordance to field development targets. This advisory system helps in maximizing the profitability and minimizing CAPEX and OPEX. This further maximizes utilization of production optimization models by 30%. Currently the system was implemented in one of ADNOC Onshore field and expected to be scaled to other fields based on consistent value creation. A hybrid approach of physics and machine learning based solution led to the development of automated workflows to identify and rank the inactive strings, well conversion to gas lift candidates & underperforming candidates resulting into successful cost optimization and production gain.

Machine-Learning Techniques Assist Data-Driven Well-Performance Optimization

2021 ◽  
Vol 73 (10) ◽  
pp. 63-64
Author(s):  
Chris Carpenter
Keyword(s):  
Machine Learning ◽  
Real Time ◽  
Performance Optimization ◽  
Production Optimization ◽  
Data Driven ◽  
Machine Learning Techniques ◽  
Well Performance ◽  
Learning Techniques ◽  
Expected Performance ◽  
Gas Lift

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 201696, “Robust Data-Driven Well-Performance Optimization Assisted by Machine-Learning Techniques for Natural-Flowing and Gas-Lift Wells in Abu Dhabi,” by Iman Al Selaiti, Carlos Mata, SPE, and Luigi Saputelli, SPE, ADNOC, et al., prepared for the 2020 SPE Annual Technical Conference and Exhibition, originally scheduled to be held in Denver, Colorado, 5–7 October. The paper has not been peer reviewed. Despite being proven to be a cost-effective surveillance initiative, remote monitoring is still not adopted in more than 60% of oil and gas fields around the world. Understanding the value of data through machine-learning (ML) techniques is the basis for establishing a robust surveillance strategy. In the complete paper, the authors develop a data-driven approach, enabled by artificial-intelligence methodologies including ML, to find an optimal operating envelope for gas-lift wells. Real-Time Well-Performance Optimization Wellsite Measurement and Control. - Flow Tests. - Past tests include sporadic measurement of multiphase rates and the associated flowing pressure and temperature, collected at various points of the production system, from bottomhole to separator conditions. Flow tests are also known as well tests; however, the authors use the term “flow test” in this paper to avoid confusion with well testing as used in pressure transient tests, including temporary shut-in pressure buildups (for producers) and pressure falloff tests (for injectors). Normally, a well would have limited data points from the past well tests (i.e., less than 50 valid flow tests in a period of 5–10 years). This data is the basis of creating ML models. Continuous Monitoring. - Every well should have adequate instrumentation, and its supporting infrastructure should include reliable power supply, minimum latency telemetry, and desktop access to production parameters. Alignment among real-time data and relational databases is required. Remote Control and Automated Actuation. - In addition to controllable valves, every well should be enabled with actuators to control the process variables. Remote control allows the operator to make changes to the current well-site configuration. Regulatory and Supervisory Control. - The value of automated closed-loop regulatory and supervisory control is to sustain optimal production while providing high well availability. Real-Time Production Optimization. - Continuous production optimization means that expected performance is challenged frequently by updating an optimal forecast with upper-level targets and current asset status. This is achieved by applying actions that close the gap between actual and expected performance. Faster surveillance loops compare actual vs. expected performance to determine minute, hourly, and daily gaps. A slower surveillance loop updates the asset’s expected performance. Well-Management Guidelines. - These are established, known limits to address and honor restrictions such as concession-contract obligations and legal limits, optimal reservoir management, flow assurance, economics, safety, and integrity.

Hybrid Electrical-Submersible-Pump/Gas-Lift Application to Improve Heavy Oil Production: From System Design to Field Optimization

2021 ◽  
pp. 1-28
Author(s):  
Son Tran ◽  
Vu Le
Keyword(s):  
Service Life ◽  
Power Consumption ◽  
Heavy Oil ◽  
Oil Production ◽  
Production Optimization ◽  
Submersible Pump ◽  
Hybrid Mode ◽  
Field Development ◽  
Electrical Submersible Pump ◽  
Gas Lift

Abstract The typical challenge encountered in developing heavy-oil reservoirs is inefficient wellbore lifting caused by complex multiphase flows. The literature on modeling of a hybrid artificial lift (AL) system is relatively sparse and these works typically model the AL system on the basis of individual AL methods. This paper presents a case study of the design and optimization of a hybrid AL system to improve heavy-oil production. We systematically design and model a hybrid electrical-submersible-pump/gas-lift (ESP/GL) system to enhance wellbore lifting and production optimization. We found that the implementation of hybrid ESP/GL system provides the flexibility to boost production and reduces production downtime. Results from the pilot test show that the production rate in hybrid mode is approximately 30% higher than in ESP-only mode. The power consumption of the hybrid mode is 3% lower in the ESP-only mode. Furthermore, the average ESP service life exceeds six years which is better than expected in the field development plan. The pump-performance-curve model is built with corrections for density and viscosity owing to the increased water production. We observed a higher pressure drawdown with GL injection at fixed ESP frequency. The GL injection reduces the density of the fluid column above the ESP, resulting in less pressure loss across the pump, less power consumption, and potentially extended service life. The nodal-analysis results suggest that the pump capacity can be considerably expanded by manipulating the GL rate instead of increasing the frequency.

Data Driven Workflow to Optimize Eagle Ford Unconventional Asset Development Plan Based on Multidisciplinary Data

2021 ◽  
Author(s):  
Tarik Abdelfattah ◽  
Ehsaan Nasir ◽  
Junjie Yang ◽  
Jamar Bynum ◽  
Alexander Klebanov ◽  
...  
Keyword(s):  
Machine Learning ◽  
Random Forest ◽  
Subject Matter ◽  
Data Driven ◽  
Well Performance ◽  
Eagle Ford ◽  
Field Development ◽  
Well Design ◽  
Machine Learning Model ◽  
Feature Importance

Abstract Unconventional reservoir development is a multidisciplinary challenge due to complicated physical system, including but not limited to complicated flow mechanism, multiple porosity system, heterogeneous subsurface rock and minerals, well interference, and fluid-rock interaction. With enough well data, physics-based models can be supplemented with data driven methods to describe a reservoir system and accurately predict well performance. This study uses a data driven approach to tackle the field development problem in the Eagle Ford Shale. A large amount of data spanning major oil and gas disciplines was collected and interrogated from around 300 wells in the area of interest. The data driven workflow consists of: Descriptive model to regress on existing wells with the selected well features and provide insight on feature importance, Predictive model to forecast well performance, and Subject matter expert driven prescriptive model to optimize future well design for well economics improvement. To evaluate initial well economics, 365 consecutive days of production oil per CAPEX dollar spent (bbl/$) was setup as the objective function. After a careful model selection, Random Forest (RF) shows the best accuracy with the given dataset, and Differential Evolution (DE) was used for optimization. Using recursive feature elimination (RFE), the final master dataset was reduced to 50 parameters to feed into the machine learning model. After hyperparameter tuning, reasonable regression accuracy was achieved by the Random Forest algorithm, where correlation coefficient (R2) for the training and test dataset was 0.83, and mean absolute error percentage (MAEP) was less than 20%. The model also reveals that the well performance is highly dependent on a good combination of variables spanning geology, drilling, completions, production and reservoir. Completion year has one of the highest feature importance, indicating the improvement of operation and design efficiency and the fluctuation of service cost. Moreover, lateral rate of penetration (ROP) was always amongst the top two important parameters most likely because it impacts the drilling cost significantly. With subject matter experts’ (SME) input, optimization using the regression model was performed in an iterative manner with the chosen parameters and using reasonable upper and lower bounds. Compared to the best existing wells in the vicinity, the optimized well design shows a potential improvement on bbl/$ by approximately 38%. This paper introduces an integrated data driven solution to optimize unconventional development strategy. Comparing to conventional analytical and numerical methods, machine learning model is able to handle large multidimensional dataset and provide actionable recommendations with a much faster turnaround. In the course of field development, the model accuracy can be dynamically improved by including more data collected from new wells.

scholarly journals Assessment of machine-learning techniques in predicting lithofluid facies logs in hydrocarbon wells

2019 ◽  
Vol 7 (3) ◽  
pp. SF1-SF13 ◽  
Author(s):  
Saba Keynejad ◽  
Marc L. Sbar ◽  
Roy A. Johnson
Keyword(s):  
Machine Learning ◽  
Oil And Gas ◽  
Hybrid Approach ◽  
Oil And Gas Industry ◽  
Added Value ◽  
Machine Learning Techniques ◽  
Superior Performance ◽  
Technical Errors ◽  
Production Stages ◽  
Log Interpretation

Wireline log interpretation is a well-exercised procedure in the oil and gas industry with all its added value from exploration to production stages. It becomes even more important when it is one of only a few available alternatives to compensate for the lack of core samples in a study of lithologic and fluid variations in a well. Yet, as with other purely expert-oriented interpretational techniques, there is always a considerable risk of subjective or technical errors. We have adopted a hybrid approach that links a machine-learning (ML) algorithm to the log interpretation procedure to solve these problems. We have applied this approach to two different hydrocarbon (HC) fields with the aim of predicting the HC-bearing units in the form of lithofluid facies logs at different well locations. The values of these logs are labels of classes that are separated based on their lithologic and fluid content characteristics. After training different MLs on the designed lithofluid facies logs, we chose a bagged-tree algorithm to predict these logs for the target wells due to its superior performance. This algorithm predicted HC units in an accurate interval (above the HC-fluid contact depth), and it showed a very low false discovery rate. The high-accuracy rate, speed of analysis, and its generalization ability, even in data-deficient cases, accentuate why including ML algorithms can improve the understanding of the subsurface at every phase of the exploration and production process. The proposed approach of using ML algorithms, trained and tuned based on the expert’s knowledge of the reservoir, can be modified and applied to future wells in a HC field to significantly minimize the risk of false HC discoveries.

Automated Geosteering Optimization Using Machine Learning

2021 ◽  
Author(s):  
Alexey Vasilievich Timonov ◽  
Rinat Alfredovich Khabibullin ◽  
Nikolay Sergeevich Gurbatov ◽  
Arturas Rimo Shabonas ◽  
Alexey Vladimirovich Zhuchkov
Keyword(s):  
Machine Learning ◽  
Horizontal Well ◽  
Target Function ◽  
Function Optimization ◽  
Geological Model ◽  
Well Performance ◽  
Learning Methods ◽  
Well Placement ◽  
Field Development ◽  
Machine Learning Methods

Abstract Geosteering is an important area and its quality determines the efficiency of formation drilling by horizontal wells, which directly affects the project NPV. This paper presents the automated geosteering optimization platform which is based on live well data. The platform implements online corrections of the geological model and forecasts well performance from the target reservoir. The system prepares recommendations of the best reservoir production interval and the direction for horizontal well placements based on reservoir performance analytics. This paper describes the stages of developing a comprehensive system using machine-learning methods, which allows multivariate calculations to refine and predict the geological model. Based on the calculations, a search for the optimal location of a horizontal well to maximize production is carried out. The approach realized in the work takes into account many factors (some specific features of geological structure, history of field development, wells interference, etc.) and can offer optimum horizontal well placement options without performing full-scale or sector hydrodynamic simulation. Machine learning methods (based on decision trees and neural networks) and target function optimization methods are used for geological model refinement and forecasting as well as for selection of optimum interval of well placement. As the result of researches we have developed the complex system including modules of data verification and preprocessing, automatic inter-well correlation, optimization and target interval selection. The system was tested while drilling hydrocarbons in the Western Siberian fields, where the developed approach showed efficiency.

scholarly journals Smart Proxy Modeling of SACROC CO2-EOR

Fluids ◽  
2019 ◽  
Vol 4 (2) ◽  
pp. 85 ◽  
Author(s):  
Gholami Vida ◽  
Mohaghegh D. Shahab ◽  
Maysami Mohammad
Keyword(s):  
Machine Learning ◽  
Oil Recovery ◽  
Simulation Models ◽  
Learning Technologies ◽  
Production Optimization ◽  
Reservoir Properties ◽  
Flow Models ◽  
Proxy Models ◽  
Co2 Eor ◽  
Proxy Modeling

Large CO2-enhanced oil recovery (EOR) projects usually contain an abundance of geological and good performance data. While this volume of data leads to robust models, it often results in difficult to manage, slow-running numerical flow models. To dramatically reduce the numerical run-times associated with the traditional simulation techniques, this work investigated the feasibility of using artificial intelligence and machine learning technologies to develop a smart proxy model of the Scurry Area Canyon Reef Operators Committee (SACROC) oilfield, located in the Permian Basin, TX, USA. Smart proxy models can be used to facilitate injection-production optimization for CO2-EOR projects. The use of a coupled grid-based, and well-based surrogate reservoir model (SRM) (also known as smart proxy modeling) was investigated as the base of the optimization. A fit-for-purpose coupled SRM, which executes in seconds, was built based on high-resolution numerical reservoir simulation models of the northern platform of the SACROC oilfield. This study is unique as it is the first application of coupled SRM at a large oilfield. The developed SRM was able to identify the dynamic reservoir properties (pressure, saturations, and component mole-fraction) at every grid-block, along with the production characteristics (pressure and rate) at each well. Recent attempts to use machine learning and pattern recognition to build proxy models have been simplistic, with limited predictive capabilities. The geological model used in this study is comprised of more than nine million grid blocks. The high correlation between the actual component and SRM, which can be visualized by mapping the properties, along with the fast footprint of the developed model demonstrate the successful application of this methodology.

The Power of Predictive Analytics in Oil Field Development: Integrating Machine Learning with Reservoir Hydrocarbon Data to Enable Enhanced Oil Recovery of Hugin Formation within the Theta Vest Structure

2021 ◽  
Author(s):  
L. T. Hardanto
Keyword(s):  
Machine Learning ◽  
Fluid Flow ◽  
Enhanced Oil Recovery ◽  
Oil Recovery ◽  
Production Optimization ◽  
Oil Field ◽  
Injection Strategy ◽  
Field Development ◽  
Facies Classification ◽  
Resolution Graph

Machine learning is an algorithm based on pattern recognition and the concept that computers can learn without being programmed to perform specific tasks. Machine learning applications that are commonly used in the oil and gas companies are petrophysical estimation and well log classification, seismic structural identification, production forecasting, and artificial intelligence tasks. The goal of this study is to integrate machine learning workflows to evaluate how reservoir hydrocarbon distribution can help prospecting, field development, and production optimization, especially 4D seismic studies. Also to observe the fluid flow and to detect bypassed oil pockets changes during the production. The workflow consists of three phases: planning, execution, and delivery. The first phase consists of collecting and preprocessing wells, seismic and interpretation data. Once the plan is considered satisfactory, it will be followed by the execution that is started with data cleaning, processing, classification, and data validation. Machine learning methods are then deployed to build an electrofacies and reservoir distribution model for the Hugin Formation using Multi-Resolution Graph-Based Clustering (MRGC). After these models reach a satisfactory level, seismic attribute analysis is performed using Principal Component Analysis (PCA) and Democratic Neural Network Association (DNNA) to create a facies probability volume. The last step in this phase is to detect geobodies of oil sand and propose an infill well or injection strategy to enable the enhancement of the oil recovery. Once the machine learning results are satisfying, tthe status of the workflow will change from execution to the delivery phase to create the final project presentation. In our study, DNNA has demonstrated excellent prediction and facies classification to image a large volume encompassing some wellbores, changes in the fluid flow during production between baseline, and monitoring seismic surveys with a good Matthews correlation coefficient of 0.849554. It allows the operator to observe the dynamic processes in and around the reservoir to help the placement of infill wells more effectively, increas development and production success, reduce risk when following proposed infill wells. The integration of machine learning can also improve the understanding of hydrocarbons in the field. It shapes E&P business strategies in a way that may increase profit revenues, such as enhanced oil recovery of an effective and efficient infill well and optimizing an injection strategy.

Integrated Asset Modeling in Mature Offshore Fields: Challenges and Successes

2014 ◽  
Author(s):  
R.. Nazarov ◽  
P.. Zalama ◽  
M.. Hernandez ◽  
C.. Rivas
Keyword(s):  
Production Management ◽  
Treatment Plant ◽  
Integrated Model ◽  
Production Optimization ◽  
Production Model ◽  
Peak Performance ◽  
Fluid Composition ◽  
Well Performance ◽  
System Pressure ◽  
Gas Lift

Abstract Production management in mature fields is a very challenging task which involves a multidisciplinary technical approach to minimize the decline rate and extend the life of the asset/field. Most of the time Integrated Asset Modeling (IAM) techniques are applied to green fields with main objectives of identifying the “bottlenecks” or to forecast production with different development cases. In the case of mature fields it is mostly considered as an optional study with less analytical value due to low operating surface pressures, already existing facilities, known well performance and studied reservoir geology. Nevertheless the processing of the reservoir, production and operational data in mature assets through one integrated workflow facilitates field management overall, thereby helping in the estimation of the remaining reserves and indicating real opportunities for optimization not seen by initial engineering scenarios. Additionally, IAM should be incorporated before getting to EOR studies. This paper describes the applied reservoir engineering workflow and integrated production model for the TSP fields (Teak, Samaan and Poui) located in the South East of Trinidad. TSP fields are jointly owned by by Repsol (70%), Petrotrin (15%) and NGC (15%) and are operated by Repsol. Current production of TSP is 13, 500 bopd. The oil produced from these fields is generally light oil, with an average range of 25-40 API and a solution GOR 200-1400scf/stb. Gas lift is the artificial lift system used in 95% of the wells. Average water cut is around 85%. Interaction of Production Engineering, Subsurface, Drilling, HSE, Facilities, and Maintenance departments is the key aspect to sustain the efficient operability of the TSP fields and operate at peak performance in spite of ageing installations, flow assurance problems and depleted reservoirs. The implementation of Operated Asset Structure in TSP in 2013 reinforced the cooperation between departments to achieve the main goals: minimum production deferrals, production optimization, screening of new opportunities and reserves, process improvement, facilities maintenance and effective logistics. Additionally, the Integrated Asset Modeling has been incorporated as part of the engineering surveillance which includes 3 fields, 100 wells, gas lift injection network, gas compressors, water treatment plant, etc. Real data from different sources and platforms, such as pressure temperature sensors, daily measured well parameters, reported operational figures, monthly welltests and screened remaining reserves are jointly transferred to the integrated model, built in commercial software (GAP/RESOLVE), bringing the field data processing and production management to the state-of-the-art level. Gas lift volume availability and system pressure, performed rigless intervention jobs (including recompletion of new zones), change of the fluid composition in certain wells, reconfiguration of facilities are timely reflected in the TSP integrated model. Based on the sensitivity runs and output results immediate actions are taken to comply with the production target.

Well GOR Prediction from Surface Gas Composition in Shale Reservoirs

2021 ◽  
Author(s):  
Alexandra Cely ◽  
Andrei Zaostrovski ◽  
Tao Yang ◽  
Knut Uleberg ◽  
Margarete Kopal
Keyword(s):  
Machine Learning ◽  
Learning Model ◽  
Production Optimization ◽  
Percentage Error ◽  
Field Development ◽  
Reservoir Fluid ◽  
Well Location ◽  
Machine Learning Model ◽  
Early Production ◽  
Shale Reservoirs

Abstract There are increased development activities in shale reservoirs with ultra-low permeability thanks to the advances in drilling and fracking technology. However, representative reservoir fluid samples are still difficult to acquire. The challenge leads to limited reservoir fluid data and large uncertainties for shale play evaluation, field development, and production optimization. In this work, we built a large unconventional reservoir fluid database with more than 2400 samples from shale reservoirs in Canada, Argentina, and the USA, comprising early production surface gas data and traditional PVT data from selected shale assets. A machine learning approach was applied to the database to predict gas to oil ratio (GOR) in shale reservoirs. To enhance regional correlations and obtain a more accurate GOR prediction, we developed a machine learning model focused on Canada shale plays data, intended for wells with limited reservoir fluid data available and located within the same region. Both surface gas compositional data and well location and are input features to this model. In addition, we developed an additional machine learning model for the objective of a generic GOR prediction model without shale dependency. The database includes Canada shale data and Argentina and USA shale data. The GOR predictions obtained from both models are good. The machine learning model circumscribed to the Canada shale reservoirs has a mean percentage error (MAPE) of 4.31. In contrast, the generic machine learning model, which includes additional data from Argentina and USA shale assets, has a MAPE of 4.86. The better accuracy of the circumscribed Canada model is due to the introduction of the geospatial well location to the model features. This study confirms that early production surface gas data can be used to predict well GOR in shale reservoirs, providing an economical alternative for the sampling challenges during early field development. Furthermore, the GOR prediction offers access to a complete set of reservoir fluid properties which assists the decision-making process for shale play evaluation, completion concept selection, and production optimization.

Sign in / Sign up

Export Citation Format

Share Document