The ability of smart well completions (SWCs) to best function as intended in the field, has led to renewed focus on smart well completion optimization opportunities. The action of inflow control valves (ICVs) is key to improved management of flux imbalance and premature production delay of unwanted fluids from contributing laterals of the intelligent wells. Difficulty with the use of ICVs includes the complexity to determine all possible combinations of valve settings (given 11 possible positions per valve on each lateral). Imposing specific wellbore pressure profiles in dual or trilateral well completions, to control production, or achieve pressure responses within ideal electrical submersible pump (ESP) operational ranges for continuous uninterrupted functioning of the smart wells is even more challenging. Real-time data acquired during SWC tests at various combinations of valve positions was modelled to determine contributions from motherbores and/or laterals in 11 smart wells. The ICV positions and surface chokes were controlled, while flow rate, wellhead pressure/temperature, and intake pressure/temperature were measured. From the tests, the optimal settings for individual or commingled flow were determined in real time from the network models (objective functions include minimizing water, maximizing oil, subject to staying within effective ESP intake pressures). Tests showed wells with up to 34% and 57% basic sediment and water (BS&W) from the motherbore and lateral respectively and corresponding normalized rates of 1.539/1.138. Implementing results of the model resulted in a normalized oil gain of 3.800 and 27% BS&W reduction from both laterals. The model yielded the optimization parameters, which resulted in choking back drainage points to achieve the water cut reduction and realize the oil gain. Test data revealed an accuracy of more than 90% match between calculated and measured flow rates of targeted drainage points. The results confirm that modeling ICV completions to determine optimum valve positions and rate targets is possible for an entire SWC or specific laterals.

In this study, a review of production performance of four existing horizontal producers equipped with Inflow Control Device (ICD) completions was conducted using 4-D dynamic modelling on a sandstone reservoir with high water mobility. The aim of this study was to investigate the optimum regulation degree across ICD completion i.e. the ratio of pressure drop across ICDs to the reservoir drawdown, suitable to delay water breakthrough, minimize water cut and achieve production balance. A single wellbore model was built by populating rock and fluid properties in 3-D around the wellbore for each of the studied wells. The model was then calibrated to the measured production log flow profile and bottomhole pressure profile for the deployed ICD completion in each well. Thereafter, several ICD simulation cases were run at target rates for a production forecast of 4 years. An optimum ICD case for each well was selected on the basis of water breakthrough delay, water cut reduction and incremental oil gain. The study results showed that there is a correlation between reservoir heterogeneity index, well productivity index (PI) and optimum regulation degree required across ICD to achieve longer water breakthrough delay and better water cut control. In general, high heterogeneity, high PI wells require higher regulation degree across ICD of close to one; medium heterogeneity, low PI require regulation degree across ICD of between 0.3 – 0.45 while low heterogeneity, low PI, require very low regulation degree of between 0.1 – 0.15. Based on study results, a new ICD design framework and correlation chart were developed. This framework was then applied to two newly drilled horizontal producers to test the applicability of the workflow in real time ICD design scenarios and positive results were achieved. Given the significant number of ICD completions deployed yearly, this new ICD design framework would provide guidance on how much pressure drop across ICD is required during real time design for newly drilled or sidetrack wells and would ultimately ensure maximum short and long term benefits are derived from deployment of ICD completions.

The subject field is located 50 miles offshore Angola in ∼1200' of water. The compliant piled tower that supports this development in Block 14 is one of the largest manmade structures in the world. The 46-well development plan includes both wet and dry trees. The multi-layered reservoirs are heterogeneous in nature. To extract the most value from this development, Chevron has developed and deployed an integrated solution that uses people, processes and technology to improve performance. In 2002 Chevron began exploring how integrated digital solutions could help better manage the reservoirs, wells and facilities. Since that time, a fully fledged i-field™ program was developed with a vision for transforming operations through technology. It's about making sure the right people have the right information at the desired frequency in a usable format at the right time to make the best decision. It's about teamwork and collaboration across function and distance. It's about providing people the time and space to develop creative alternatives and solutions to effectively manage the daily and longer term challenges. The Production Optimization Solution for the subject field was successfully deployed in September 2010. The solution is designed to improve production engineering productivity, increase production and recovery by optimizing overall field management, and decrease response time through early problem detection. Two initial engineering workflows were deployed along with an asset overview dashboard that displays key information, trends and results. After a rigorous evaluation process, a project team was created to develop the engineering workflows that make use of nodal analysis and pressure transient analysis tools for field and model management. A 3rd party vendor was selected to design and manage the data and application integration required to successfully implement the solution. This paper will detail the development and deployment of this integrated solution that is allowing the engineers to spend more time on technical work, respond quicker to anomalies, and make better, faster decisions. It will also discuss the requirements for success that include change management, training, support and organizational capability.

Smart/Intelligent well completions are broadly used to maximize multilateral well productivity, restrict unwanted water and gas production, and improve sweep efficiency. To achieve the optimum economic values of smart completions, the surface and subsurface chock valves settings need to be frequently optimized using the best-in-class techniques. Applying the right optimization technique will ensure a successful and efficient optimization. The paper discusses an innovative production optimization approach using real-time modeling based on nodal analysis for multilateral wells. These multilateral wells are equipped with surface and subsurface downhole valves with various choke settings and downhole permanent pressure gauges. The technique utilizes the data collected during a conventional optimization and a commercial steady-state model. It estimates the flowing parameters of individual laterals, determines the optimum pressure drop across each downhole valve, and estimates productivity of each lateral during the commingled production at various choke valves settings. The approach was successfully field-tested and validated. The generated models were used to predict well performance at various conditions. The approach starts by collecting well rates and flowing bottom-hole pressure data at various chokes settings including commingled and individual lateral testing. The acquired data are used to calibrate the model, generate different production scenarios and optimize the performance of each lateral. Adoption of the technique among others facilitate better management of the multilateral wells production to fulfil both short and long term objectives, namely, optimizing production of these wells and improving recovery. In addition to the reduction of OPEX associated with the conventional procedures to test and optimize these wells.

In this study, we aim to demonstrate how machine learning can empower computational models that can predict the flow rate of a given well. Given current real-time data and periodic well tests, this new method computes flow rates using data-driven model. The computational model is based on analyzing the relations and trends in historical data. Relational databases include huge amounts of data that have been accumulated throughout decades. In addition, there is a large number of incoming operational data points every second that gives a lot of insight about the current status, performance, and health of many wells. The project aims to utilize this data to predict the flow rate of a given well. A variety of well attributes serve as inputs to the computational models that find the current flow rate. Artificial Neural Networks (ANN) were used in order to build these computational models. In addition, a grid search algorithm was used to fine-tune the parameters for the ANN for every single well. Building a single unique model for every well yielded the most accurate results. Wells that are data-rich performed better than wells with insufficient data. To further enhance the accuracy of the models, models are retrained after every incoming patch of real-time data. This retraining calibrates the models to constantly represent the true well performance and predict better. In practice, Flow rate prediction is used by production engineers to analyze the performance of a given well and to accelerate the process of well test verification. One of the main challenges in building unique models for every well is fine-tuning the parameters for the artificial neural networks, which can be a computationally intensive task. Parameter fine-tuning hasn't been discussed in previous literature regarding flow rate prediction. Therefore, our unique approach addresses the individuality of every well and builds models accordingly. This high-level of customization addresses the problem of under-fitting in ANN well models.