Hardware Acceleration of High-Performance Computational Flow Dynamics Using High-Bandwidth Memory-Enabled Field-Programmable Gate Arrays

Scientific computing is at the core of many High-Performance Computing applications, including computational flow dynamics. Because of the utmost importance to simulate increasingly larger computational models, hardware acceleration is receiving increased attention due to its potential to maximize the performance of scientific computing. Field-Programmable Gate Arrays could accelerate scientific computing because of the possibility to fully customize the memory hierarchy important in irregular applications such as iterative linear solvers. In this article, we study the potential of using Field-Programmable Gate Arrays in High-Performance Computing because of the rapid advances in reconfigurable hardware, such as the increase in on-chip memory size, increasing number of logic cells, and the integration of High-Bandwidth Memories on board. To perform this study, we propose a novel Sparse Matrix-Vector multiplication unit and an ILU0 preconditioner tightly integrated with a BiCGStab solver kernel. We integrate the developed preconditioned iterative solver in Flow from the Open Porous Media project, a state-of-the-art open source reservoir simulator. Finally, we perform a thorough evaluation of the FPGA solver kernel in both stand-alone mode and integrated in the reservoir simulator, using the NORNE field, a real-world case reservoir model using a grid with more than 10 5 cells and using three unknowns per cell.

Finding the Best of the Best: Hydraulic Fracturing Design Optimization Study in Oman

Development of the tight gas Khazzan Field in Sultanate of Oman has progressed through an extensive learning curve over many years. Thereby, the hydraulic fracturing design was fine-tuned and optimized to properly fit the requirements of the challenging Barik reservoir in this area. In 2018, BP Oman started developing the Barik reservoir in the Ghazeer Field, which naturally extends the reservoir boundary south of Khazzan Field. However, the Barik reservoir in the Ghazeer area is thicker and more permeable than in the Khazzan Field; therefore, the hydraulic fracturing design required adjustment to be optimized to directly reflect the reservoir needs of the Ghazeer Field. A comprehensive hydraulic fracturing design software was used for this optimization study and sensitivity analysis. This software is a plug-in to a benchmark exploration and production software platform and provides a complete fracturing optimization loop from hydraulic fracturing design sensitivity modelled with a calibrated mechanical earth model to detailed production prediction using the incorporated reservoir simulator. One of the stimulated wells from Ghazeer Field was used as the reference for this study. The reservoir sector model was created and adjusted to match actual data from this well. The data include fracturing treatment execution response, surveillance data such as radioactive tracers, bottomhole pressure gauge, and pressure transient analysis. Reservoir properties were also adjusted to match long-term production data obtained for this reference well. After the reservoir model was fully validated against actual data, multiple completion and fracturing scenarios were simulated to estimate potential production gain and thus find an optimal hydraulic fracturing design for Ghazeer Field. Many valuable outcomes can be concluded from this study. The optimal treatment design was identified. The value of fracture half-length versus conductivity was clarified for this area. The comparison between single-stage fracturing versus multistage treatment across the thick laminated Barik reservoir in a conventional vertical well was derived. The drainage of different layers with variable reservoir properties was compared for a range of different scenarios.

Field-Scale Modeling of Smart Completion Tools for Optimum Recovery

Flow control devices (FCDs) like inflow control devices (ICDs) and interval control valves (ICVs) (i.e., equalizer) have increased applications in both conventional and unconventional resources. They have been used to mitigate water or gas coning problems for mature fields in conventional reservoirs, to alleviate premature water breakthrough in naturally fractured reservoirs, and to optimize the steam distribution in heavy oil reservoirs. There have been increased trend in using FCDs in the real field. Previously, complex well models have been implemented in a large-scale parallel reservoir simulator by Tareq et al. (2017). The implementation can simulate an intelligent field contains tens to hundreds of multilateral complex wells commonly referred in the literature as maximum reservoir contact (MRC) wells with mechanical components such as ICVs and ICDs. In this paper, a new framework to model controlling the FCDs in complex well applications will be presented. The implementation is integrated into a complex well model. It can be easily used to model the dynamical control of devices. Simulation studies using both sector model and field model have been conducted. A systematic full-field operation is used for device control applications of smart wells. Successful application of field level controls in smart wells has the benefit of the improved overall GOSP performance.

Thermally Controlled Fluid to Reduce Formation Breakdown Pressures in Tight Gas Reservoirs

High breakdown pressure is one of the major challenges in deep tight gas reservoirs. In certain wells, achieving breakdown pressures within the completion tubular yield limit is not possible, and those zones may have to be abandoned without fracturing. Using thermally controlled fluid can lower the formation temperature and ultimately reduce the stresses of the tight gas reservoir formation near the wellbore. The objective of this study is to prove numerically that having a cooled near-wellbore region is a feasible and effective solution to reduce the breakdown pressure. An integrated hydraulic fracturing and reservoir simulator that has been developed at the University of Texas at Austin is utilized for this study. The simulator is a non-isothermal, multi-phase black-oil flow in reservoir, fracture, and wellbore domains. It was found that using thermally controlled fluid is effective in reducing breakdown pressure. Bottomhole Pressure (BHP) decreased by up to around 60% when the temperature of the near-wellbore region is reduced by 60 °F under the simulated conditions in this study. Injecting thermally controlled fluid did not only reduce the high breakdown pressure but also improve the hydraulic fractures efficiency and complexity. This technique is novel and has not been studied in depth in the literature. Utilizing thermally controlled fluid can be a cost effective solution to reduce high breakdown pressure in tight gas reservoirs.

Core-to-Field-Scale Simulations of Polymer Viscoelastic Effect on Oil Recovery Using the Extended Viscoelastic Model

Being one of the most commonly used chemical EOR methods, polymer flooding can substantially improve both macroscopic and microscopic recovery efficiencies by sweeping bypassed oil and mobilizing residual oil, respectively. However, a proper estimation of incremental oil to polymer flooding requires an accurate prediction of the complex rheological response of polymers. In this paper, a novel viscoelastic model that comprehensively analyzes the polymer rheology in porous media is used in a reservoir simulator to predict the recovery efficiency to polymer flooding at both core- and field-scales. The extended viscoelastic model can capture polymer Newtonian and non-Newtonian behavior, as well as mechanical degradation that may take place at ultimate shear rates. The rheological model was implemented in an open- source reservoir simulator. In addition, the effect of polymer viscoelasticity on displacement efficiency was also captured through trapping number. The calculation of trapping number and corresponding residual-phase saturation was verified against a commercial simulator. Core-scale tertiary polymer flooding predictions revealed the positive effect of injection rate and polymer concentration on oil displacement efficiency. It was found that high polymer concentration (>2000 ppm) is needed to displace residual oil at reservoir rate as opposed to near injector well rate. On the other hand, field-scale predictions of polymer flooding were performed in a quarter 5-spot well pattern, using rock and fluid properties representing the Middle East carbonate reservoirs. The field-simulation studies showed that tertiary polymer flooding might improve both volumetric sweep efficiency and displacement efficiency. For this case study, incremental oil recovery by polymer flooding is estimated at around 11 %OOIP, which includes about 4 %OOIP residual oil mobilized by viscoelastic polymers. Furthermore, the effect of different parameters on the polymer flooding efficiency was investigated through sensitivity analysis. This study provides more insight into the robustness of the extended viscoelastic model as well as its effect on polymer injectivity and related oil recovery at both core- and field-scales. The proposed polymer viscoelastic model can be easily implemented into any commercial reservoir simulator for representative field-scale predictions of polymer flooding.

Intelligent Waterflood Optimization Advisory System – A Step Change Towards Digital Transformation

In this paper, we discuss the development of an end-to-end waterflood optimization solution that provides monitoring and surveillance dashboards with artificial intelligence (AI) and machine learning (ML) components to generate and assess insights into waterflood operational efficiency in an automated manner. The solution allows for fast screening of waterflood performance at diverse levels (reservoir, sector, pattern, well) enabling prompt identification of opportunities for immediate uptake into an opportunity management process and for evaluation in AI-driven production forecast solution and/or a reservoir simulator. The process starts with the integration of a wide range of production and reservoir engineering data types from multiple sources. Following this, a series of monitoring and surveillance dashboards of key units and elements of the entire waterflood operations are created. The workflows in these dashboards are framed with key waterflood reservoir and production engineering concepts in mind. The optimization opportunity insights are then extracted using automated traditional and AI/ML algorithms. The identified opportunities are consolidated in an optimization action list. This list is passed to an AI-driven production forecast solution and/or a reservoir simulator to assess the impact of each scenario. The system is designed to improve the business-time decision-making cycle, resulting in increased operational performance and lower waterflood operating costs by consolidating end-to-end optimization workflows in one platform. It incorporates both surface and subsurface aspects of the waterflood and provides a comprehensive understanding of waterflood operations from top-down field, reservoir, sector, pattern and well levels. Its AI/ML components facilitate understanding of producer-injector relationships, injector dynamic performance, underperformance of patterns in the sector as well as evaluating the impact of different optimization scenarios on incremental oil production. The data-driven production forecast component consists of several ML models and is tailored to assess their impact on oil production of different scenarios such as changes in voidage replacement ratio (VRR) in reservoir, sector, pattern and well levels. Opportunities are also converted into reservoir simulator compatible format in an automated manner to assess the impact of different scenarios using more rigorous numerical methods. The scenarios that yield the highest impact are passed to the field operations team for execution. The solution is expected to serve as a benchmark, upon successful implementation, for optimizing injection schemas in any field or reservoir. The novelty of the system lies in automating the insights generation process, in addition to integrating with an AI/ML production forecasting solution and/or a reservoir simulator to assess different optimization scenarios. It is an end-to-end solution for waterflood optimization because of the integration of various components that allow for the identification and assessment of opportunities all in one environment.

Modeling a Novel Approach to Delay the Water Breakthrough in Gas Cap Wells Using Smart Completions: Case Study Onshore Abu Dhabi Field

This paper presents modeling a novel approach to determine the impact of implementing smart completions on water injectors located near the periphery of the gas cap and on gas producing wells situated in the gas cap of a giant Middle East onshore field. The objective of the study is to thoroughly investigate different smart completion designs which can effectively delay water breakthrough on the gas cap wells. The study investigates the impact of adding smart well completion designs like ICD and AICD valves in delaying water breakthrough. The first phase involves adding smart completions to only water injectors. Sensitivity runs on several downhole completion design scenarios are conducted using a commercial near wellbore simulator and the optimal downhole completion design is implemented on a dynamic model and its impact is examined using a reservoir simulator. In the second phase, this approach is applied only for gas producers, and in the third phase the smart completions are simultaneously applied to both water injectors and gas producers. The detailed study has revealed that the uncertainties and time involved in selecting optimal ICD design and placements could be reduced considerably by using an optimized workflow. The workflow uses a carefully designed process of using the outcomes from near wellbore simulators and incorporating the results in the actual full field dynamic models to assess the field level impacts. When compared to the bare foot design, ICD and AICD valves showed better performance in delaying water breakthrough from the gas wells. This paper provides a detailed study on the impact of different smart completions on delaying water breakthrough in gas production wells. The study also investigates how a uniform injection or production profile can be produced using different smart completions. Uniform injection and production profiles limit water fingering in the reservoir, and thereby delay water breakthrough caused by the flow of water through high permeability channels.

High Resolution Reservoir Simulator Driven Custom Scripts as the Enabler for Solving Reservoir to Surface Network Coupling Challenges

This paper presents a practical methodology of optimizing and building a detailed field surface network system by using the high-resolution reservoir simulator driven custom-made Python scripts to efficiently predict the future performance of the vast oil and gas-condensate carbonate field. All existing surface hydraulic tables are quality checked and lifting issue constraints corrected. Pressure losses at the wellhead chokes incorporated into the high-resolution reservoir simulator in the form of equation by using the custom scripts instead of a table format to calculate gas rate dependent pressure losses more precisely. Consequently, all 400+ surface production system manifolds, pipes and well chokes Horizontal Flow Performance (HFP) tables are updated and coupled to the reservoir simulator through Field Management (FM) controller which in turn generates Inflow Performance Relationship (IPR) tables for the coupled wells and passes them to solve the network. The methodology described in this paper applied for a complex field development planning of the Karachaganak. At present, reservoir management strategy requires constant balancing effort to uniformly spread gas re-injection into the lower Voidage Replacement Ratio areas in the Upper Gas-Condensate part of the reservoir due to reservoir heterogeneity. Additionally, an increase in field and wells gas-oil ratio and water-cut creates bottlenecks in the surface gathering system and requires robust solutions to decongest the surface network. Current simulation tools are not always effective due longer run times and simulation instability due to complex network system. As a solution, project-specific network balancing challenges are resolved by incorporating custom-made scripts into the high-resolution simulator. Faster and flexible integrated model based on hydraulic tables reproduced the historical pressure losses of the surface pipelines at similar resolution and generated accurate prediction profiles in a twice-quicker time than existing reservoir simulator. Overall, this approach helped to generate more stable production profiles by identifying bottlenecks in the surface network and evaluate future projects with more confidence by achieving a significant CAPEX cost savings. The comprehensive guidelines provided in this paper can aid reservoir modeling by setting up flexible integrated models to account for surface network effects. The value of incorporating Python scripts demonstrated to implement non-standard and project specific network balancing solutions leveraging on the flexibility and the openness of the modelling tool.

Reservoir Simulation of CO2 Storage Using Compositional Flow Model for Geological Formations in Frio Field and Precaspian Basin

CO2 storage is a greenhouse gas mitigation instrument for many countries. In this paper, we investigate the possibility of CO2 storage in the region of the Precaspian basin using the compositional flow model that was verified by the data of the Frio pilot project, USA. We use local grid refinement in the commercial reservoir simulator. In the reservoir simulation for data of the Frio Pilot project, we have achieved a good history matching of well pressure. Different scenarios were tested, and post-injection migration was shown for both case studies. The long-term reservoir simulation shows the potential amount of trapped CO2 by residual and dissolved trapping mechanisms in the Precaspian basin. The performed uncertainty study covered the uncertainty of the model’s parameters resulting in P10, P50 and P90 cases in terms of the amount of trapped CO2.

Seawater Breakthrough Monitoring and Reservoir-Model Improvement Using Natural Boron

Natural geochemical data, which refer to the natural ion concentration in produced water, contain important reservoir information, but is seldomly exploited. Some ions were used as conservative tracers to obtain better knowledge of reservoir. However, using only conservative ions can limit the application of geochemical data as most ions are nonconservative and can either interact with formation rock or react with other ions. Besides, mistakenly using nonconservative ion as being conservative may cause unexpected results. In order to further explore the nonconservative natural geochemical information, the interaction between ion and rock matrix is integrated into the reservoir simulator to describe the nonconservative ion transport in porous media. Boron, which is a promising nonconservative ion, is used to demonstrate the application of nonconservative ion. Based on the new model, the boron concentration data together with water production rate and oil production rate are assimilated through ensemble smoother multiple data assimilation (ES-MDA) algorithm to improve the reservoir model. Results indicate that including nonconservative ion data in the history matching process not only yield additional improvement in permeability field, but also can predict the distribution of clay content, which can promote the accuracy of using boron data to determine injection water breakthrough percentage. However, mistakenly regarding nonconservative ion being conservative in the history matching workflow can deteriorate the accuracy of reservoir model.