The SPE has split the former "Management & Information" technical discipline into two new technical discplines:
- Data Science & Engineering Analytics
The SPE has split the former "Management & Information" technical discipline into two new technical discplines:
|Theme||Visible||Selectable||Appearance||Zoom Range (now: 0)|
Abstract Reservoir simulators based on physics provide the most accurate method for predicting oil and gas recovery, in particular from waterflood and EOR processes. However, detailed full-field simulation can be computationally demanding. In recent years, there have been attempts in accelerating reservoir simulation by combining simplification of the gridding requirement with data-driven approaches while maintaining the full physics. One such approach is the physics-based data-driven flow network model where 1D or 2D grids connecting the wells are configured and simulated. The parameters of the flow network model are then tuned to match full 3D simulation or field-data. Even though the grid has been simplified, a large number of parameters are needed to reproduce the 3D simulation results. In this paper, an approach similar to the flow network model is presented. The main contribution of this paper is the parameterization of the gridding process between the wells such that a minimal number of parameters are needed. Essentially, the grids between the wells are configured to model accurately the flow behavior. The corner-point grid geometry is kept so that current simulators could be used with the proposed method. In this paper, the grid geometry is determined with AI methods for one waterflood run. The grid could be used subsequently for waterflood with widely different injection/production scenarios and even for chemical flood. The ability of the approach to derive the grid from a single waterflood run is another significant contribution of this paper.
Al-Hajri, Hamood S. (Petroleum Development Oman) | Al-Sawafi, Marwan (Petroleum Development Oman) | Al-Hashimi, Abdulaziz R. (Petroleum Development Oman) | Al-Hadidi, Khalsa (Petroleum Development Oman) | Al-Kindi, Osama M. (Petroleum Development Oman) | Al-Amri, Mohammed (Petroleum Development Oman) | Al-Abri, Mohammed (Petroleum Development Oman) | Al-Hinai, Suleiman (Petroleum Development Oman)
Abstract Water and chemical EOR are the main secondary recovery mechanisms in many heavy oil fields in Oman. The development concept during EOR phase is through intense infill drilling with narrow well spacing. Field-M is currently under secondary recovery phase with both water and chemical EOR (Polymer) development. During this phase, water production increases significantly and all undesired water is being disposed through disposal wells. This increases carbon intensity as disposal process generates CO2 emissions with no additional benefit, which considered as uneconomical emissions. Due to increased amount of produced water during this phase, water handling capacity (including water disposal) was fully utilized to maximize oil production from this field. Creative solutions were certainly needed reduce uneconomical water disposal and increase oil gain. As per the field development, certain pre-defined polymer dosage need to be mixed with treated produced water to achieve a viscosity of around 15 cp to ensure effectiveness of chemical EOR. Field-M injection strategy was suggested to be under controlled fracture condition to maximize throughput. In controlled fracture injection environment, monitoring fracture propagation is very important as it can cause direct interference with producers leading to injection fluid short circuiting. Fracture propagation can be determined using pressure fall off test. In addition, water quality must be monitored regularly as it plays a major role in fracture propagation. Effective surveillance and sampling plan was generated and implemented to ensure to ensure effectiveness of the polymer injection and to capture any opportunities related to increasing injection within the field. The analytical work showed that fracture propagation is a function of injection pressure, injection rate, fluid properties (in this case produced water quality and polymer quality) and in-situ stresses. Most of this parameters are controls though effective surveillance, metering & sampling. However in-situ stress condition is dynamic as the reservoir pressure keeps changing based on dynamic changes in injection and offtake. Thus, fracture propagation was monitored carefully through periodic temperature surveys and pressure fall off test to identify opportunities to optimize injection in some of the injectors. The findings from these activities enabled increasing injection rate up to 30% in some of the injection patterns. This optimization provided additional sink for the produced water reducing water disposal and uneconomical CO2 emissions by at least 5%. This is considered this as the first step toward zero water disposal goal. In addition increasing injection in these patterns resulted in significant increase in oil gain associated with polymer injection peaking to maximum of 42% in some of the injector/producers patterns. The effective use of surveillance data was key enabler to achieve ultimate goal of increasing polymer injection and reduce carbon intensity within the field. This goal was achieved with significant gain of oil.
Nguyen, Quang Minh (The University of Tulsa) | Onur, Mustafa (The University of Tulsa) | Alpak, Faruk Omer (Shell International Exploration & Production Inc.)
Abstract Summary This study focuses on carbon capture, utilization, and sequestration (CCUS) via the means of nonlinearly constrained production optimization workflow for a CO2-EOR process, in which both the net present value (NPV) and the net present carbon tax credits (NPCTC) are bi-objectively maximized, with the emphasis on the consideration of injection bottomhole pressure (IBHP) constraints on the injectors, in addition to field liquid production rate (FLPR) and field water production rate (FLWR), to ensure the integrity of the formation and to prevent any potential damage during life-cycle injection/production process. The main optimization framework used in this work is a lexicographic method based on line-search sequential quadratic programming (LS-SQP) coupled with stochastic simplex approximate gradients (StoSAG). We demonstrate the performance of the optimization algorithm and results in a field-scale realistic problem, simulated using a commercial compositional reservoir simulator. Results show that the workflow is capable of solving the single-objective and bi-objective optimization problems computationally efficiently and effectively, especially in handling and honoring nonlinear state constraints imposed onto the problem. Various numerical settings have been experimented with to estimate the Pareto front for the bi-objective optimization problem, showing the trade-off between the two objectives NPV and NPCTC. We also perform a single-objective optimization on the total life-cycle cash flow, which is the aggregated quantity of NPV and NPCTC, and quantify the results to further emphasize the necessity of performing bi-objective production optimization, especially when utilized in conjunction with commercial flow simulators that lack the capability of computing adjoint-based gradients.
Tariq, Zeeshan (King Abdullah University of Science and Technology) | Gudala, Manojkumar (King Abdullah University of Science and Technology) | Yan, Bicheng (King Abdullah University of Science and Technology) | Sun, Shuyu (King Abdullah University of Science and Technology) | Rui, Zhenghua (China University of Petroleum, Beijing)
Abstract The potential for large-scale storage of carbon dioxide (CO2) through Geological Carbon Sequestration (GCS) in deep geological formations such as saline aquifers and depleted oil and gas reservoirs is significant. Effectively implementing GCS requires evaluating the risk of plume confinement and storage capacity at each site through a thorough assessment. To assess the stability of the caprock after CO2 injection, efficient tools are needed to evaluate the safe duration of CO2 injection. This study used Particle Swarm Optimization (PSO) evolutionary algorithm to optimize the maximum CO2 storage capacity in saline aquifers without risking the integrity of the caprock. A deep learning (DL) model, fully connected neural networks, was trained to predict the safe injection duration. The movement of CO2 was simulated for 170 years following a 30-year injection period into a deep saline aquifer using a physics-based numerical reservoir simulator. The simulation took into consideration uncertainty variables such as petrophysical properties and reservoir physical parameters, as well as operational decisions like injection rate and perforation depth. Sampling the reservoir model with the Latin-Hypercube approach accounted for a range of parameters. Over 720 reservoir simulations were performed to generate training, testing, and validation datasets, and the best DNN model was selected after multiple executions. The three-layer FCNN model with 30 neurons in each layer showed excellent prediction efficiency with a coefficient of determination factor over 0.98 and an average absolute Percentage Error (AAPE) less than 1%. The trained models showed a good match between simulated and predicted results and were 300 times more computationally efficient. PSO was utilized to optimize the operational parameters in the DL models to achieve maximum CO2 storage with minimum damage to the caprock. The results suggest that the DNN-based model can serve as a reliable alternative to numerical simulation for estimating CO2 performance in the subsurface and monitoring storage potential in GCS projects.
de Oliveira, Josias Pereira (University of Campinas) | Santos, Susana Margarida da Graça (University of Campinas) | dos Santos, Antônio Alberto Souza (University of Campinas) | Schiozer, Denis José (University of Campinas)
Abstract Many projects in the Brazilian pre-salt assume the use of water alternating gas (WAG-CO2) injection as an ecologically safe carbon storage strategy, with improved hydrocarbon recovery. However, studies that compare these advantages with a simpler management plan are not common. The objective of this work is to compare WAG-CO2 injection with continuous injection of water and gas (CIWG) rich in CO2 in separate wells for the development and management of a light-oil fractured carbonate reservoir subject to full gas recycling. We employed the UNISIM-II benchmark model, a naturally fractured carbonate reservoir with Brazilian pre-salt characteristics, which enables an application in controlled environment where the reference response is known (UNISIM-II-R). We used a model-based decision analysis for production strategy selection, hierarchical optimization of the decision variables and algorithms to maximize the objective function. Representative models (RM) are selected from the ensemble of models and used to incorporate the effects of geological, reservoir, and operational uncertainties into the optimization process. The net present value is the objective function during the nominal optimization of candidate strategies of each RM and the expected monetary value and risk analysis are considered to select the final production strategy considering uncertainties. The risk analysis was quantified based on downside risk and upside potential relation to a benchmark return. We optimized two alternative development plans (one considering WAG-CO2 injection and the other continuous injection of water and gas in separate wells) and compared their performance indicators and decision variables, including design variables (number, type and placement of well, and size of production facilities) and life-cycle control rules (management of equipment over time). We then applied a cross-simulation, where the best strategy optimized for one recovery method was applied to the other and the injection strategy was optimized again. We were therefore able to assess the need to pre-define the recovery method before defining design variables to validate the flexibility of each strategy for possible future changes in the recovery mechanism. Finally, we repeated the study for different reservoir scenarios to compare the alternatives considering typical uncertainties of the Brazilian pre-salt and validated the final strategies in the reference model to quantify the real value in decision making. The strategies reached a full gas recycling in both recovery methods and allowed a comparison of their advantages and disadvantages. The operations of WAG-CO2 injection can be more complex and the equipment more expensive. The novelty of this work is the consideration of continuous injection of water and gas in separate wells as a simpler alternative to the development and management of pre-salt oil fields, since this method may also meet operators’ and environmental demands, bearing simpler operating challenges and promoting good recovery and profitability.
Abstract The present work introduces an efficient workflow for AI-enhanced decision-making in Field Development Planning Optimization. Despite the clear importance of uncertainty quantification in decision-making, we find that constraints in time, hardware, and costs are often limiting factors during field evaluation, with the drawback of having a biased uncertainty description or a wrong risk perception. The proposed work encompasses history matching, solution analysis, and production optimization with special emphasis on reducing both simulation and processing time, maximizing what we can call the result per core hour. At the center of our work is an AI-guided optimizer suited to avoid excessive convergence bias and maintain an optimal exploration vs. exploitation performance. The optimizer allows the integration of a multi-objective (MO) formulation in standard history matching and optimization workflows. Despite the flexibility of MO optimization and the vast literature in the energy industry, its usage in real-field cases has always been quite limited due to its formulation availability in commercial software and the increased computation time. This work will show improvement in solution accuracy and formulation flexibility compared to Single Objective (SO) formulations at no increase in runtime. MO is based on the iterative convergence of an efficient frontier from the results generated by the simulation. This same concept has been brought to a user analysis step to allow the identification of best solutions across multiple evaluation workflows, lowering the expertise level for a solution.
Abstract In this work, we develop computationally efficient methods for deterministic production optimization under nonlinear constraints using a kernel-based machine learning method where the cost function is the net present value (NPV). We use the least-squares support-vector regression (LSSVR) to maximize the NPV function. To achieve computational efficiency, we generate a set of output values of the NPV and nonlinear constraint functions, which are field liquid production rate (FLPR) and water production rate (FWPR) in this study, by running the high-fidelity simulator for a broad set of input design variables (well controls) and then using the collection of input/output data to train LS-SVR proxy models to replace the high-fidelity simulator to compute NPV and nonlinear constraint functions during iterations of sequential quadratic programming (SQP). To obtain improved (higher) estimated optimal NPV values, we use the existing so-called iterative sampling refinement (ISR) method to update the LSSVR proxy so that the updated proxy remains predictive toward promising regions of search space during the optimization. Direct and indirect ways of constructing LSSVR-based NPVs as well as different combinations of input data, including nonlinear state constraints and/or the bottomhole pressures (BHPs) and water injection rates, are tested as feature space. The results obtained from our proposed LS-SVR-based optimization methods are compared with those obtained from our in-house StoSAG-based line-search SQP programming (LS-SQP-StoSAG) algorithm using directly a high-fidelity simulator to compute the gradients with StoSAG for the Brugge reservoir model. The results show that nonlinear constrained optimization with the LSSVR ISR with SQP is computationally an order of magnitude more efficient than LS-SQP-StoSAG. In addition, the results show that constructing NPV indirectly using the field liquid and water rates for a waterflooding problem where inputs come from LSSVR proxies of the nonlinear state constraints requires significantly fewer training samples than the method constructing NPV directly from the NPVs computed from a high-fidelity simulator. To the best of our knowledge, this is the first study that shows the means of efficient use of a kernel-based machine learning method based on the predictor information alone to perform efficiently life-cycle production optimization with nonlinear state constraints.
Yan, Bicheng (King Abdullah University of Science and Technology) | Gudala, Manojkumar (King Abdullah University of Science and Technology) | Sun, Shuyu (King Abdullah University of Science and Technology)
Abstract Geothermal energy is naturally renewable harnessed from subsurface reservoirs and is feasible to help enrich the energy spectrum and decarbonize the economy. Cold geo-fluid such as water is injected to extract the heat from hot rocks, and then hot fluid can be produced, which can be used for the purpose of heating or power generation. The reservoir management of geothermal recovery process is an integration of geology, drilling, reservoir and production, and particularly it requires expensive simulations that couple the thermo-hydro-mechanical (THM) effect. In this study, we developed a reservoir simulation model to simulate the enhanced geothermal systems (EGS). After evaluating the produced fluid temperature curves, we proposed a generalized thermal decline model that considers the thermal breakthrough and the following decline behavior. This model is parsimonious with only 3 variables. Moreover, a forward surrogate model by deep neural network is developed to predict the decline model variables and the ultimate total net power based on the reservoir parameters. The forward surrogate is integrated with a differential evolution optimizer, which considers reservoir uncertainties and nonlinear constraints for the optimization of the total net power. Accelerated by the thermal decline model and forward surrogate model, we were able to efficiently perform reservoir optimization in high-performance computing environment, and this makes the workflow quite scalable for real-time reservoir management.
Zhang, Yi (State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, Shanghai) | Ma, Ning (State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, Shanghai) | Gu, Xiechong (State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, Shanghai) | Shi, QiQi (State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, Shanghai)
_ This paper introduces a method combining Bayesian optimization (BO) with multilayer perceptron (MLP) neural network regression to build a surrogate model in the process of automatic ship hull form optimization. The MLP regression replaces computer fluid dynamics solvers for solving computationally expensive numerical simulations. The MLP model has many parameters called hyperparameters, which largely determine the prediction accuracy of the model; BO tracks the prior distribution to obtain the optimal hyperparameters. By training the data collected through Latin hypercube sampling, the MLP model composed of optimized hyperparameters has higher regression accuracy. A Korea Research Institute of Ships & Ocean Engineering container ship was used as the verification model. The results show that this procedure is more effective and consumes less time in ship hull optimization. Introduction Ship hull form design has an important influence on ship hydrodynamic performance. Simulation-based design (SBD) technology opens up a new situation for ship hull form design. This technology can help design engineers to explore the design space under constraints and automatically get the optimal design plan (Campana, 2006; Cheng et al., 2018; Harries et al., 2019; Yu et al., 2019; Ichinose and Taniguchi, 2022). An automated SBD needs to integrate a parametric molder, a hydrodynamic performance solver, and an optimization algorithm.
Jong, Siaw Chuan (Hess Exploration and Production Malaysia B.V., Kuala Lumpur, Malaysia) | Aziz, Khairil Faiz Abdul (Hess Exploration and Production Malaysia B.V., Kuala Lumpur, Malaysia) | Goo, Jia Jun (Hess Exploration and Production Malaysia B.V., Kuala Lumpur, Malaysia) | Hiew, Ronnie (Hess Exploration and Production Malaysia B.V., Kuala Lumpur, Malaysia) | Strickland, Kenny (Hess Exploration and Production Malaysia B.V., Kuala Lumpur, Malaysia) | Hussin, Arief (Hess Exploration and Production Malaysia B.V., Kuala Lumpur, Malaysia) | Yusof, Khazimad (Hess Exploration and Production Malaysia B.V., Kuala Lumpur, Malaysia) | Macleod, Andy (Hess Exploration and Production Malaysia B.V., Kuala Lumpur, Malaysia) | Yusoff, Syukur (Hess Exploration and Production Malaysia B.V., Kuala Lumpur, Malaysia) | Chung, Chay Yoeng (Hess Exploration and Production Malaysia B.V., Kuala Lumpur, Malaysia) | Liew, Alex (Hess Exploration and Production Malaysia B.V., Kuala Lumpur, Malaysia)
Abstract High temperature, high carbon dioxide coupled with hydrogen sulfide contents, and rapid PPFG pressure ramp increase gas development well tends to cause high well capex for Operator. This well type typically needs high CRA material with at least a 10,000-psi rated system to complete. Offshore peninsular Malaysia’s North Malay Basin (NMB)’s deep reservoirs also fall into the described category. This paper aims to share the optimization journey, applications, and learnings of the company’s H.T. sour-rated 10Ksi gas development wells through several phases, besides fulfilling the gas delivery need for the country. In addition, engineering and operational optimizations are identified to reduce the well’s time and cost without sacrificing the crew’s safety as the team focus. The company wells engineering team applied Lean approaches encompassing the complete Plan-Do-Check-Adjust cycle to achieve the optimization. Well data usage, lessons learned, collaboration, continuity, and striving for continuous improvements are the key factors to ensure good optimization results. Fit-for-purpose drilling and completions equipment design and application, rig offline capabilities planning, wellhead dummy hanger plug design for offline cementing, intervention-less production packer setting device, offline annulus nitrogen cushion fluids displacement and other applications will be explained in the paper. The paper explained the operational challenges, how and what optimizations applied to achieve excellent well performance compared to targets and previous campaigns. The wells team optimizations spread out from engineering to execution stages, including rolling out in-house talent of digitization and digitalization of well performance surveillance, in line with the industry's way forward. The recent campaign post optimization concluded with no safety incidents, below budgeted time and cost, low overall NPTs, and achieved first gas to meet the country's power generation demand. Open, collaborative, and proactive cross departments communications are the catalysts that contributed to the positive optimization journey's results.