Search

Advanced or smart completion wells are different from conventional wells by being equipped with downhole flow control devices such as Interval Control Valves (ICV) and Inflow Control Devices (ICD) to offer improved reservoir management and control and thus maximise hydrocarbon production and recovery. In order to justify their implementation and increase their economic return, a high degree of robustness in modelling, prediction and optimisation of their performance is required. To improve the robustness of forecasting production from advanced or smart wells using reservoir simulation, high-level details in rock and fluid flow properties are needed in the near-wellbore region to accurately capture the flow dynamics. The paper presents an improved approach that enables us to robustly predict the performance of advanced or smart wells in reservoir simulation and highlights the importance of representing the near-wellbore region when optimizing smart well completions. Performances of advanced or smart well completions are very dependent on changes in flow rate, pressure, and saturations, which mainly occur in areas around the wells. The paper demonstrates the use of local grid refinement (LGR) in the near-wellbore region to enhance the accuracy level of simulation predictions. In the study, an objective function based ICV optimization strategy was used to identify the optimum settings for every time step during the simulation run. We also demonstrate how to correlate ICV settings to Passive Inflow Control Device (PICD) or Autonomous Inflow Control Device (AICD) strengths if a requirement arises to impose the use of ICVs. Using a well-established synthetic reservoir model, we demonstrate how the representation of the near-wellbore region impacts reservoir performance predictions and influences the way ICVs and ICDs are optimized. We observe that by applying this approach, the predicted NPV and recovery factor change by 6.6% and 6.1%, respectively. In addition, this study also quantifies the impact of near-wellbore representation on four completion types; Openhole, ICV, PICD and AICD completions. The novelty of this paper is that it presents an approach to improve production forecasts that supports decision making during field development planning to maximize profit and minimize risk.

In this work, we implement new software for improved waterflood management by combining classical finite volume reservoir simulation together with streamline tracing and corresponding inter- well flux evaluations to optimize waterflood performance. We show reduced water cut and improved oil recovery by using streamline-based flux information to adjust well rates, while retaining the advantages of rigorous finite volume simulation. We have introduced two basic modules here: a commercial reservoir simulator Eclipse and our own streamline tracing and waterflood management program. Waterflood simulation is performed for a certain time span until simulation is paused and the streamline tracing program is called to calculate inter-well fluxes and adjust new well rates for better waterflood performance. The simulation continues afterwards until the next tracing and adjustment point is reached. The two modules work iteratively. The streamline tracing program uses the classical Pollock's tracing method. It is designed to trace streamlines on a compressible velocity field and a general corner point grid system with non- neighboring connections. The new injection rates are adjusted according to each well's injection efficiency calculated from inter-well fluxes as proposed in previous works. Our waterflood management program successfully works cooperatively with the commercial simulator, with data transfer realized by making the simulator to output the velocity field and other information into restart files. Streamline tracing is performed successfully not only on simple geometry corner point grid cases, but also on heavily-faulted realistic reservoirs under waterflood. Streamline- based inter-well multi-phase fluxes are calculated, in which the injection efficiencies are derived accordingly. Injectors with higher efficiencies will receive more water from the source, according to the injection allocation algorithms and vice versa. After re-adjusting injection rates multiple times during the simulation, we typically observe a reduction in field water cut of up to 5% and an increase in oil recovery in our test cases. Inter-well flux information serves as effective diagnostic tools to identify injector-producer pairs with large amount of water cycling. All simulations conducted here are rigorously finite volume based, which takes into account the full physics of non-advective processes such as gravity and capillary effects. In conclusion, we have implemented a streamline-based waterflood management program which works iteratively and cooperatively with a commercial reservoir simulator, without switching to streamline simulation. It provides an effective solution for improving oil recovery in brown fields by combining the rigorous mathematical nature of finite volume simulation and the power of streamline- based flood management.

Creating reservoir model that is accurate requires extensive data and enormous computing resources. As these resources are often not available, this study proposes a workflow (& corresponding case study) to generate box reservoir models that require minimum data yet yield reasonable results. Field Lima has two wells Lima-1 & Lima-2 (down-dip well) that produces gas from Sand-B which is a moderate water drive reservoir. After producing gas naturally, Lima-2 watered-out while Lima-1 still produced gas. To curtail the increase of Water Gas Ratio (WGR) in Lima-1, a coproduction technique was evaluated. The intent was to produce water from Lima-2 at high rates to reduce WGR increase in Lima-1. This required reservoir simulation to predict if producing water from Lima-2 will limit aquifer encroachment at Lima-1 and justify the additional CAPEX required to install water handling facility at Lima-2. As Lima is a marginal field, available dataset only included; initial reservoir pressure, production history and standard open-hole logs. No permeability, aquifer or pressure history data was available. Therefore, a workflow was formulated to create a reservoir model that can predict production rates of Lima-1 using available data within engineering accuracy. Firstly, the pressure history was created using recorded flowing wellhead pressures and wellbore hydraulics model. Further, a tank model was created for the reservoir using material balance software and cumulative productions of both wells were used as unique data to calibrate the model. This was used to calculate the aquifer properties. Fractional flow models & Corey correlations were used to estimate relative permeability data. All this derived data was then used as input in the reservoir simulator, where a localized block grid was created using depth-structure maps of Sand-B. The model was history matched using production data of Lima-1 & Lima-2 and aquifer movement was calibrated using water breakthrough in Lima-2. Finally, the model was run to predict production rates of Lima-1 with & without keeping Lima-2 on production to evaluate if co-production technique was effective. The model predicted that Lima-1 will be able to deliver its production targets even without keeping Lima-2 on production suggesting that extra CAPEX required for co-production was not justified. Forecasted gas rates for Lima-1 matched the actual gas rates with around 10% error. The accuracy of forecasted results endorses the use of box reservoir models to undertake economic decisions, in scenarios where lack of time & data doesn’t allow creation of detailed dynamic reservoir models.

Reservoir studies, including preparation of field development plan, are processes typically dominated by time constraints. In general, reservoir studies consist in multiple geoscience activities integrated to build a fine geological model that eventually leads to an upscaled numerical model suitable for history matching and forecast simulations. In the simulation stage, the quality and effectiveness of the activity is highly dependent on the computational efficiency of the numerical model. This is particularly true for complex, supergiant carbonate reservoirs. Often, even with today's simulators, upscaling is still needed and simplifications can be implemented to allow computationally intensive history matching and risk analysis workflows. This paper provides some real field examples where these issues were faced and successfully managed, without applying further simplifications to the geological concept of the model: principles of reservoir simulations and common sense reservoir engineering were used to adjust properties of the model and then speed-up numerical simulation. Tuning included a combination of various solutions, such as deactivating critical cells whenever possible, calibrating convergence and time stepping control, tweaking field management to prevent instability in the computation, optimization of number of cores and cells split among cores to optimize load balancing and scalability. These solutions were used on two super-giant carbonate fields, a triple porosity (matrix, karst and fractures) undersaturated light oil reservoir and a supercritical gas and condensate reservoir. The former field was described using an upscaled model of about 700,000 active cells and a dual porosity - dual permeability formulation; the latter was described by a relatively coarse model of about 400 thousand active cells using a single porosity formulation. Large speed-up, up to five times with respect to reference simulations, was achieved without simplifying the geology and losing accuracy perceivably. Benefits were achieved for both conventional and high-resolution simulators.

In traditional Discrete Fracture Network (DFN) method for fractured porous media simulation, fractures are modelled in different scales and orientations using single computational grid. However, a large number of small grids near both ends of fracture are generated, which substantially reduce the computation efficiency. In this paper, a new line fracture approach is presented to avoid generation of small grids. Combined with Multi-Point Flux Approximation (MPFA) algorithm, the simulation efficiency is highly improved without losing accuracy. We present a modified mesh generation algorithm in which the volume of large fractures is set zero by neglecting the fracture width. The fracture aperture (very short line) is treated as single node during mesh generation using the paving method. Therefore, the number of small grids around the fracture is significantly decreased. In this paper, a quadrilateral unstructured grid system is generated using the proposed paving method. Then, by taking the median dual of quadrilateral mesh as the control volume, all fracture properties such as the aperture and permeability are accounted for as independent parameters in discrete fractures. The governing equations of DFN model are derived and solved using MPFA algorithm for oil-water two-phase flow. A new discrete fracture network model using line fracture approach is developed. To show the efficiency and accuracy of the new model, the proposed method is applied to a case study of two separated and two crossed fractures, in the computation domain with the scale of 50 by 50 meters. Compared with the standard DFN model including the width of fractures, the computation efficiency of the new model is highly improved (by 2.5 times), due to the large reduction of grid numbers (20%). Good agreement in water and oil saturation distributions resulted from the two models validates the newly developed method. Simulation of naturally fractured reservoirs has been a challenge for decades. Discrete fracture network modelling is a promising approach. However, without an efficient grid generation method, DFN model is not feasible and practical due to the large amount of computational time. The novelty of the present work is the improved model using line fracture approach. The new grid generation algorithm is coupled with multi-point flux approximation approach for flow simulation in fracture network system. High efficiency and accuracy of the developed method is readily applied to fractured reservoir simulation. Sponsored by NSFC (Natural Science Fund of China - "the numerical simulation of natural fractured reservoir based on the unstructured hexahedral grid system and concurrent computation") (No. 51474179)

A workflow has been developed for a shale play in North America for the strategic deployment of refracturing in an unconventional reservoir. Although it is clearly understood in the industry that refracturing in tight unconventional reservoirs can have a dramatic impact on the productivity and longevity of economic production, until now there has been no workflow that could be counted on to deliver consistent results. The impact of production and stress re-orientation plays a significant role in adding a fourth dimension of time to the existing three-dimensional problem of hydraulic fracture modeling in these reservoirs.We examined the re-orientation of reservoir stress and the change in stress magnitude with production in a case from the Eagle Ford. The stress redistribution occurring due to poroelastic effects was simulated using a finite element geomechanical simulator and then tied back to predict the new hydraulic fracture growth. Fracture geometry as it relates to draining the unstimulated domain was found to play a significant role in fracturing success. The impact of adding new perforations in the wellbore and application of a suitable diversion technique were also considered in the study. Our numerical investigation of the potential increase in productivity with selection of the optimum refracturing treatment timing and technique led to development of a methodology that solves the complex four-dimensional problem.By adopting the guidelines from the case study application of the end-to-end workflow, operators can avoid the futility of too early or too late fracturing and thereby maximize their return on investment.

Reservoir simulation is an important tool for petroleum engineers to predict oil production and optimize the management of oil fields. Nowadays, large-scaled reservoir simulations are required by oil industry to simulate complex geological models in order to obtain high resolution results. When the reservoir engineers design new production processes, it is necessary to run dozens of simulations to find optimal solutions. The speed of serial reservoir simulators can be a big challenge. Parallel reservoir simulators with fast computational methods should be invested. In this paper, we developed cost-effective parallel reservoir simulation techniques, which include mathematic fluid models, FIM discretization method, multilevel preconditioners and its implementations on shared memory. Based on the mathematical characteristics of the pressure and saturation in black-oil model, a multilevel preconditioner is set up which includes algebraic multigrid method (AMG), incomplete LU factorization, et al. The preconditioner is implemented on share memory to speed up the efficiency of numerical simulation. Several numerical experiments such as the benchmark test two-phase and three-phase SPE10 simulation and a field-scale mature waterflooding reservoir simulation were performed to test the efficiency, robustness of the proposed simulation technique and its parallel speedup. In the numerical experiments, the proposed reservoir simulation techniques can successfully simulate numerical cases with different reservoir properties on a desktop computer. Numerical results show that the proposed reservoir simulation techniques are quite efficient in solving the nonlinear PDEs and robust for highly heterogeneous field-scale problems.

Waterflood induced fractures are the new features affecting the reservoir heterogeneity in low-permeability reservoirs. Due to the high injection pressure, waterflood induced fractures will generate and propagate when the formation pressure exceeds fracturing pressure during the waterflood process of low-permeability reservoirs. Waterflood induced fractures will interact with natural fractures and create high permeable flow channels. Conventional reservoir simulation models often fail in simulating the waterflooding of low-permeability reservoirs without considering the effect of reservoir pressure on the fractures. Current fracture models based on Perkins-Kern-Nordgren (PKN) model only focus on the simulation of the propagation of a single fracture without considering the interaction between waterflood induced fractures and pre-existing natural fractures. Therefore, based on the analysis of field data and the mechanism of waterflood induced fractures, a comprehensive wateflood induced fracture simulation model for naturally fractured low-permeability reservoirs is established. Compared with the existing models, this waterflood induced fracture model can simulate the process of occurrence, propagation and closing of waterflood induced fractures due to the effect of high injection pressure and complex in-situ stress. The model considers the interaction between induced fractures and natural fractures, including slippage and crossing. The fracture model is coded in a module of a reservoir simulator. The simulation results shows that waterflood induced fractures worsen the reservoir heterogeneity, greatly increase the water cut of production well and lower the oil production rate. The model can simulate the impact of waterflood induced fractures on waterflooding and optimize well patterns in low-permeability reservoirs to improve sweep efficiency and oil recovery.

This paper describes the integrated approach to decision making on the optimized horizontal well completion with multistage fracturing (MSF) based on petrophysical, geomechanical and numerical modeling. This approach is pertinent in case of insufficient target exploration including absence of a full scale geo&hydromodel (FSGHM). The basic work process of the proposed approach is based on consistent efforts of a petrophysicist, geomechanical engineer, fracturing engineer and reservoir engineer and thus includes petrophysical and geomechanical modeling, fracturing design, numerical modeling and final planned well flow rate evaluation. The role of geomechanics in the production chain is to determine zones with the lowest and the highest fracture gradients to control fracture location and analyze fracture geometry. The basic objective of the proposed approach is to develop a set of recommendations to select: Optimized frac sleeve spacing to increase production during a certain period of well production;Optimized proppant injection volume in general and for each fracture;Recommended horizontal well paths for efficient multistage fracturing completion and reduction of the most probable risks;Frac sleeve placement with provision for the horizontal well path through the section (vertically) to optimize fracture initiation points and maximize reservoir coverage. Besides, according to the base history, costs of new target development using available production technologies can be estimated based on the completed work package. An additional objective of this approach is identification of interdependent parameters for further facilitation of downhole surveys and laboratory studies and excluding of non-informative methods. This paper gives an example of the developed method application for a prospecting and appraisal well at one of the West Siberian fields with a target characterized by extremely low permeability and uncertainty.

Nowadays there is observed a global trend of ever-increasing complexity and heterogeneity of hydrocarbon reservoirs. More challenging time makes companies take high-risk decisions based on long-term production prediction. As deterministic approach always gives a single result, it loses its relevance when we deal with risks. Generally probabilistic assessment methodology is coming to the forefront. A new approach called RE-Status (Reservoir evaluation status) has been developed in this study. The methodology is based on probabilistic assessment and along with normal uncertainty technique it includes lithofacies modeling. RE-Status is based on six key parameters that allow us to understand reservoir complexity and level of project definition in terms of reservoir engineering. Geological models were created by combining all geological and geophysical data in order to provide the highest-quality 3D representation of heterogeneous carbonate reservoirs. Evaluation of micro-macro heterogeneities and complex analysis of all core studies were used to derive lithofacies. Systematization of stratigraphic analysis, core descriptions and well logging data allowed us to locate facies in reservoir cross-section. It gave us better understanding of the reservoir and nature of filtration processes. The new approach was applied to two gas fields. Impact of reservoir parameters and lithofacies variability on the dynamics of gas production was described. In the first case RE-Status helped to improve history matching by using facies and uncertainty analysis. In the second case (green field) production prediction based on the results of heterogeneity analysis and facies modeling allowed to locate sweet zones, optimize drilling strategy, increase reliability of forecasting and reduce the uncertainty range. Novelty of the described approach is application of facies modeling to reduce key uncertainties, match the reservoir model to historical data and provide the most accurate forecast of production.

In oil and gas industry, quick decisions on reservoir management have a huge impact on business success. Reservoir simulation is used as a typical tool to predict field performance and analyze uncertainties for assistance on decision making. Nevertheless, history matching, as a critical step of reservoir simulation, typically requires running a numerical simulation model repeatedly with different parameter settings, which is a huge computational cost, especially for complicated geological models with numerous grid cells. For reservoir engineers, how to achieve efficient reservoir simulation by taking full advantage of field data without compromise on the simulation time is a big concern.In this work, Smart Proxy, as a relative new proxy model type, is proposed to investigate the feasibility of fastening history matching process as an alternative. Smart Proxy is an ensemble of Artificial Intelligence and Data Mining (AI&DM) technologies that are able to learn and replicate the behavior of reservoir simulation model with high accuracy. It can be developed off line and put online for automatic history matching at high speed such that a single run can be performed in a fraction of a second (Mohaghegh 2006).This paper presents the Smart Proxy generation and its implementation in a real oilfield simulation case named SACROC Unit. It essentially involves detailing numerical reservoir simulation and Smart Proxy generation for a naturally fractured carbonate numerical simulation model with highly complicated development stages. The developed Smart Proxy is implemented to perform automatic history matching in the designated study area of SACROC Unit. The efficient history matching has been proven to be successfully accomplished using Smart Proxy simulation. Tremendous time and efforts have been saved without any compromise on simulation accuracy compared with that of traditional numerical reservoir simulation method.

This article presents the development of the UNISIM-I-M benchmark case, a set of simulation models representing the field and the information required for a decision-making analysis in the oil field management phase, giving continuity to UNISIM-I benchmark studies. Using the feedback from previous benchmark cases (UNISIM-I-D, UNISIM-I-R), we present a proposal specifically for the management phase, i.e., the period after strategy implementation. UNISIM-I-M is designed for research activities in decision analysis to select exploitation strategies during the reservoir's management phase. This work presents the results of the construction of this benchmark and information on how it can be used. UNISIM-I-M considers an exploitation strategy already in place in the initial stage of field development and a water-flooding project with 25 wells with history production data for the first 7 years. Uncertainties include structural models, PVT, oil-water contact, water relative permeability, vertical permeability multiplier, rock compressibility, 500 images with petrophysical characteristics related to facies, net-to-gross ratio, porosity, horizontal and vertical permeabilities, systems availability, and economic scenarios. The simulation model, with affordable simulation time, is based on a reference model reflecting the characteristics of Namorado Field, Campos Basin in Brazil, comprising real and synthetic data from the appraisal phase. Based on a geological model, we used a Discretized Latin Hypercube with geostatistical realizations to sample and combine different types of uncertainties (continuous and discrete attributes and realizations represented by geostatistical images) to generate 2000 possible scenarios. We then applied a probabilistic history-matching process to reduce the number of scenarios. The misfit between the models and the production history data was evaluated by quantification and diagnostic procedures, considering acceptance levels and several objective functions simultaneously, including well-fluid rates and bottom-hole pressures. This yielded a set of 48 scenarios honoring the history data and considering all uncertainties to compose UNISIM-I-M. This benchmark proposes the optimization of future design and/or control variables of the exploitation strategy for the remaining 23 years, such as infill drilling, recompletion, well conversion, conditions for shutting wells, and intelligent valves. To develop the benchmark, the studies should consider deterministic and probabilistic approaches using reservoir simulation, economic and risk analysis. All main probabilistic parameters are specified to allow a complete evaluation. We present one solution for validation. To conclude, UNISIM-I-M enables tests, application, validation of existing and new methodologies and a comparison of the results of research institutes, universities and oil companies regarding optimization and risk analysis techniques. UNISIM-I-M takes on special importance due to the lack of benchmark cases with known answers and complete reservoir models comprising geological, operational and economic scenarios. It provides challenges to test techniques for decision analysis in the management phase to the uncertainties, complexity and constraints involved and the nature of the design and control variables. The dataset is publicly available via webpage. Participants are encouraged to discuss methods, results and challenges to validate different procedures for the selection of the management strategy. To enable a fair comparison, a set of results and data with complete documentation of assumptions should be reported.

M1 reservoir is a large multi-layered sandstone reservoir of deltaic environment with oil bearing area around 1500Km2. Sand bodies of M1 is influenced by fluvial, tide and wave, which results in great variations in sand bodies distribution, reservoir quality, vertical and lateral connectivity, etc. Furthermore, M1 is still under primary depletion and current reservoir pressure is around saturation pressure. So how to successfully conduct waterflooding and effectively develop this kinds of reservoirs are very important. Through detailed geological characterization and dynamic description, 4 reservoir types with different reservoir quality and connectivity are defined and classified. Then 3 equal-probability geological models for M1 are built, which stands for Upside model with good lateral and vertical connectivity, Expected model with medium connectivity and Downside model with poor connectivity. And optimized waterflooding development strategy for different reservoir types with different stacked styles are studied through reservoir simulation based on 3 sector models. The optimized development strategy includes pressure maintenance level, well pattern, well spacing, and perforation strategy. The results show that the optimized well pattern and well spacing for different reservoir types with different connectivity expectations are different. For reservoir Type A of good reservoir quality, 1000m well spacing with inversed 9-spot pattern is optimum for Upside, Expected and Downside models. But for Type C of poor reservoir quality, 1000m well spacing with inversed 9-spot pattern is optimum for Upside model, while 700m well spacing with 5-spot pattern is optimum for Expected model, and 500m well spacing with 5-spot pattern is optimum for Downside model. For M1 reservoir, because good quality reservoir Type A and Type B are mainly developed in the South part, 1000m well spacing with 5-spot pattern is recommended. While poor quality reservoir Type C and Type D are mainly developed in the North part of M1, 700m well spacing with 5-spot pattern is recommended for this area. For perforation strategy, all zones perforated for both producers and injectors are the optimum for 1000m inversed 9-spot pattern, and only Type C and Type D layers are perforated for the infilling producers of 700m well spacing pattern. This paper offers a waterflooding development strategy case study for deltaic sandstone reservoirs, which details the optimized development strategy for different reservoir types with different stacked styles and connectivity expectations. It also provides a methodology and reference case for engineers to develop other similar fields.

This case study presents an analysis of production and pressure data from the Horst reservoir, an uplifted fault block with a thin oil rim and large gas cap in the greater Angostura field offshore Trinidad and Tobago. The study resulted in reinterpretation of the volumetrics and improved understanding of connectivity and compartmentalization in the field. This paper demonstrates how the application of fundamental Reservoir Engineering principles provides an effective way to understand reservoir flow dynamics and can guide the use of more sophisticated numerical techniques. The insights gained from this approach illustrate the importance of acquiring good quality data during the appraisal, start-up and production phases of an oil and gas field. Also, this study reinforces the benefits of using idle wells as observation wells for obtaining reservoir surveillance data. The greater Angostura field consists of Oligocene age turbidite sandstone reservoirs with multiple compartments and complex geology. Some uncertainties remain in seismic interpretation with regards to mapping reservoir boundaries due to poor seismic imaging in the area. Consequently, analysis of production performance was critical to distinguish between different scenarios that were developed using seismic volume interpretation. Formation pressures in the Horst reservoir were obtained after two years of production from offset reservoirs. Pressures were significantly different from the offset reservoirs indicating that Horst reservoir was potentially a separate compartment. The Horst commenced production after two years of offset production and was initially constrained due to high GOR. In the recent years, the Horst reservoir production has been considerably increased due to additional drilling and the commencement of gas sales from the greater Angostura field. This paper presents the results of material balance and numerical simulation studies on the strong production performance seen in the Horst reservoir. The reservoir pressures for all blocks were estimated and the potential communication between reservoirs was investigated with material balance analysis. This led to an estimate of the Horst reservoir in-place volumes that were significantly higher than the originally mapped volumes. The insight from the material balance analysis was incorporated into a numerical simulation modeling effort. A robust simulation history match was obtained with a unique but practical approach to model the pressure depletion seen in the Horst reservoir during periods of no production. The simulation model was used to develop an optimized reservoir management strategy which has been implemented to increase production and recovery from the Angostura field. Furthermore, the improved understanding of reservoir performance has led to refinements of the geological and seismic interpretation of the Horst area.

Commonly when fields are operated, there are several challenges to face and make profitable the operations, from the reservoir perspective one of the key challenges is when we have undesirable fluids that can cause cost in the operations and make more complex the production. This is the case when we have water or gas production in an oil reservoir or water production of a gas reservoir. First of all the productivity on the well can be affected, the recovery factor, and all the impact that can cause in the operations, costs, etc. When we talk about water production there are several possible sources associated to this problem, in the literature (Fig.1) there are at least ten know possible sources can cause water production, one of these problems considered is the coning phenomenon (Fig. 2). Eventually, the water or gas breaks through into the well and water from the aquifer and/or gas from the gas cap is produced along with oil. The water or gas production increases progressively after breakthrough time and may reduce significantly the crude oil production. Understand all the parameters that affect or influence coning is very important and also to be able to select the proper methods to analyze and mitigate it. In this paper the water or gas coning is examined covering some analytical methods and numerical reservoir simulation workflow applied to different reservoirs of the Veracruz Basin to understand the behavior of the wells, estimate the critical rates, the breakthrough time and establish a proper plan to this wells to avoid or mitigate the water and gas production; to reach this a numerical simulation workflow is proposed, examples and comparison is included when the wells are above or below critical rates.

XXX unit of Huabei oilfield, a complex fault-block and high viscosity reservoir, has stepped into middle period of development. Though the water cut has reached 81%, only 4.7% of the original oil in place (OOIP) has been produced, indicating that the current development method for this block may be inappropriate. In order to improve the production performance of this block, we studied the proper development technique on the basis of its geologic features. We examined the applicability of combining individual layers, and determined the upper limit for the segments of an injector as well as the optimal injection rate for each segment. In addition, we estimated the reasonable well density, the oil production rate, the injection-production ratio, and formation pressure using both numerical simulation and reservoir engineering method. We further coupled reservoir simulation with an optimization algorithm to provide proper measures for the improvement of oil recovery. The first measure examined is to divide layering segments and complete the well pattern. Our simulations give the optimized position of new wells. The primary oil recovery of two layering segments increases by 2.33% and 1.75%, respectively. The second one is the optimization of liquid production rate. In comparison with the current development state, the optimized schedule will increase the primary oil recovery of the two segments by 2.1% and 1.2%. Therefore, the optimized technique may serve as an effective approach to develop other fault-block reservoirs. In this work, several development methods demonstrate a significant effect to enhance oil recovery for complex fault-block reservoir, which can give a good guidance for other similar reservoirs.

Polymer flooding can be a viable alternative to enhance oil recovery of heavy oil fields. The comparison of water and polymer flooding requires a reliable reservoir numerical model to safely ensure the decision analysis including uncertainties and involving a significant amount of simulation runs. Reservoir engineers usually work-around this problem using fast models to reduce computational time; however, it is important to maintain the reliability of heterogeneous reservoir models. Therefore, the proposed work demonstrates how the grid size impacts on accuracy of results and both the time and computational resources. In addition, it is performed an analysis to establish how the typical physical phenomena related to the polymer flooding impact the applicability of the proposed methodology.A highly heterogeneous heavy oil field is used in our proposed studies. The application is conducted to assist pilot test wells. Initially, the definition of the reference strategy (OTM) is presented. In this case study, the OTM strategy is based on the optimization of location and number of wells. Secondly, the effect of grid size on the process is evaluated through two procedures: (1) DA method: a selection of grid size defined by drainage area of the selected wells; and (2) UP method: upscaling is performed in the simulation model. Then, production parameters resulting from models of both procedures (DA and UP) are compared with OTM results. The definition of which procedure is better to represent the wells of interest is performed. Finally, the impact of three polymer flooding physical phenomena (degradation, adsorption and Non-Newtonian effect) on production results is presented, in order to verify the accuracy of the generated fast models in maintained considering these effects.The DA method present good accuracy for reservoir response compared to the reference OTM results. Differently, UP models are not as accurate as DA models for production results indicators. In conclusion, the application of the methodology allows choosing the DA over the UP technique to support our decision in the investigated recovery methods, showing a reliable and fast procedure. Finally, this procedure shows excellent applicability and robustness when a typical polymer flooding study is carried out: the evaluation of physical phenomena impact on production results based on wells of interest.

An increasing number of field development projects include rigorous uncertainty quantification workflows based on parameterized subsurface uncertainties. Reservoir model calibration workflows for reservoir simulation models including historical production data, also called history matching, deliver non-unique solutions and remain technically challenging. In addition, the validation process of the reservoir simulation model often introduces a break of the conceptual connection to the geological model. This raises questions on how to quantify the deviation between the calibrated simulation model and the original geological model. Workflow designs for history matching require scalable and efficient optimization techniques to address project needs. Derivative-free techniques like Markov Chain Monte Carlo (MCMC) are used for optimization and uncertainty quantification. Adjoint techniques derive analytical sensitivities directly from the flow equations. For history matching those sensitivities are efficiently used for property updates on grid block level. Both techniques have different characteristics and support alternative history matching strategies like global vs. local, stochastic vs. deterministic. In this work both techniques will be applied in an integrated workflow design to the Norne field. The Norne field is a North Sea oil-and-gas reservoir with approximately 30 wells, with one third being used for WAG injection for pressure support. Field data was previously released by Statoil and made available for a public benchmark study (NTNU Norway) testing history matching techniques including production and time-lapsed seismic data. We focus on well production data for history matching. MCMC is used for global parameter updates and uncertainty quantification in a Bayesian context. An implementation of an adjoint technique is applied for analytical sensitivity calculations and local parameter adjustments of rock properties. History matching results are presented for field wide and well-by-well production data. Consistency checks between updated and original geological model are presented for rock property distribution maps. Geostatistical measures including spatial correlations are used to quantify deviations between updated and original geological model. In conclusion scalability and performance efficiency of the practical workflow implementation is discussed with a perspective of a consistent feedback loop from history matching to geological modeling.

In this work, we validate experimentally a dimensionless capillary pressure function for imbibition at mixed-wet conditionsthat we developed recently based on pore-scale modeling in rock images. The difference to Leverett's traditional J-function is that our dimensionless function accounts for wettability and initial water saturation after primary drainage through area-averaged, effective contact angles that depend on the wetting property and distribution of oil- and water-wet grain surfaces. We hypothesized that describing these effective contact angles as a function of Amott wettability index relates the pore- and core-scale wettability, and that pore-scale imbibition simulations can provide these relations for specific rocks. In the present work, we adopt the dimensionless function to scale imbibition capillary pressure data measured on mixed-wet sandstone and chalk cores. The measured data practically collapse to a unique curve when subjected to the dimensionless capillary pressure function. For each rock material, we employ the average, dimensionless curve to reproduce the measured capillary pressure curves and obtain excellent agreement. We also demonstrate two approaches to generate different capillary pressure curves at other mixed-wettability states than that available from the data used to generate the dimensionless curve. The first approach changes the shape of the spontaneous- and forced-imbibition segments of the capillary pressure curve while the saturation at zero capillary pressure is constant. The second approach shifts the vertical level of the entire capillary pressure curve, such that the Amott wetting index (and the saturation at zero capillary pressure) changes accordingly. Thus, integrating these two approaches with the dimensionless function yields increased flexibility to account for different mixed-wettability states. The validated dimensionless function scales mixed-wet capillary pressure curves from core samples accurately, which demonstrates its applicability to describe variations of wettability and permeability with capillary pressure in reservoir simulation models. This allows for improved utilization of core experiments in predicting reservoir performance. Reservoir simulation models can also employ the dimensionless function together with existing capillary pressure correlations.

This paper presents an optimization approach using a reservoir field scale compositional flow model to co-optimize oil production and CO2 storage performance in the Farnsworth field unit (FWU), Ochiltree County, Texas. A geocellular model constructed from geophysical geological and engineering data acquired from the unit was used for the study. An initial history match of primary and secondary recovery constructed was used as basis for CO2 flood study. A scenario based prediction model constructed for FWU (Ampomah et al. 2016a) was used as the baseline case for comparison to study the effect of recycling produced gas, infill drilling and water alternating gas (WAG) cycles on oil recovery and CO2 storage. A multi objective function that considers both oil recovery and CO2 storage was defined. Initial sensitivity analysis using a Latin-hypercube sampling technique was used to study the effects of operational uncertain variables on a defined objective function. A number of these operational variables were selected as control variables to be included in the optimization process. A neural network optimization algorithm was utilized to optimize the objective function both with and without uncertainty. Vertical permeability anisotropy ratio (Kv/Kh) was selected as the uncertain parameter in optimization, with uncertainty based on experience from the history match. A risk aversion factor was used to represent results at various confidence levels. The simulation results were compared to a baseline case that predicted a CO2 storage of 75% purchased CO2 and oil recovery of 72% original oil in place. The results showed an improved approach for optimizing oil recovery and CO2 storage within FWU. At the end of 21 years of prediction out of the optimal case, more than 94% of purchased CO2 has been stored and nearly 80% of the oil recovered. The sensitivity analysis reduced the number of control variables to lessen computational time. The defined objective function proved to be a robust approach to co-optimize oil recovery and CO2 storage. The Farnsworth CO2 project will serve as a benchmark for future CO2–EOR or CCUS projects in the Anadarko basin or geologically similar basins throughout the world.

SummaryFrom a technical point of view, the success or failure of LSW (low salinity waterflooding) projects strongly depends on reservoir geology; however, this has not been systematically evaluated in the past and the effects of clay minerals are often neglected in conventional reservoir simulation. This paper presents one of the first studies on integrated modeling, assisted history matching (HM) and production forecasting of field-scale LSW.To handle this complex recovery process, we used a comprehensive ion-exchange model, fully coupled with geochemistry specially designed for the modeling of LSW physical phenomena in an EOS reservoir simulator. The model is capable of accounting for the critical role of the properties, quantity, and distributions of clay minerals. We developed an integrated modeling approach that involves the combination of geological software, a reservoir simulator, and a robust optimizer in a big-loop workflow for sensitivity analysis, HM, optimization, and uncertainty assessment. The numerical simulation results indicate that LSW's performance depends critically on the reservoir geological characteristics. Multiple geological realizations can be automatically generated from the big-loop approach that are needed for fast and accurate HM and optimization of LSW. In sandstone reservoirs, clay content varies across regions, resulting in differences in ion-exchange capacity and weights of the relative permeability modification. We introduce the scaled-equivalent-fraction ion exchange which is associated with the calculated Cation-Exchange-Capacity function; the wettability alteration will be shifted based on both the ion exchange and the clay content in each grid block.The key parameters for successful field-scale LSW HM include: clay distribution/quantity associated with different facies, relative permeability modification, wettability alteration thresholds, reservoir minerals, geochemical reactions, and operating conditions. Finally, LSW HM by tuning reservoir parameters only may lead to poor prediction results, while the integrated modeling approach provides much better forecasting results to the true history data.The work presented in this paper contributes to an understanding of the critical roles of reservoir geology on the field-scale LSW performance, in particular, for substantially reducing HM errors, accurately predicting the future production, maximizing oil recovery and minimizing the risks of LSW implementation.

Reservoir models for fractured reservoirs are mainly dual-porosity models, and assume a unique equivalent fracture scale. However, fractured reservoirs are generally characterized by a multitude of fracture scales, ranging from small-scale diffuse fractures to large-scale seismic faults. This paper presents a methodology to extend single-fracture-scale dual-porosity models to situations with more scales, and investigates the optimal number of fracture scales. We focus on what happens at local level (zero-dimensional system of one cell) in case of a singlephase sequential flow through different fracture levels with some fracture pattern assumptions. Different options to turn a continuous fracture distribution into a given number of discrete levels are investigated. Eventually, based on dimensionless coefficients, an optimal number of fracture scales is determined by lumping levels with similar behaviour. Rate-limiting transfer occurs between these lumped levels. The dimensionless parameters that are defined are similar to the dimensionless parameter lambda used in dual porosity well test interpretation. They are a dimensionless representation of transfer rates. Each rate is also characterized by a dimensionless time, which is the reciprocal of the dimensionless transfer coefficient. The lumping of scales is thus done in between groups that are separated by low dimensionless transfer coefficients (slow transfer). Examples are shown of cases in which a 5-scale representation can be lumped to two scales (dual porosity) or three scales (triple porosity fracture- fracture-matrix). This study is a first step towards a reservoir modelling representing more than one single fracture scale, and sets a general methodology for lumping scales in a single-phase context. However for practical application the multiscale issue is even more relevant for multi-phase flow, and must be addressed. Such extension will be part of future work, which will also include implementing multiple scale models in 3D reservoir simulation.

The isogeometric analysis method (IGA) has been an emerging technique for exact geometrical discretization and efficient high order approximations in the field of computational engineering. While its potential has been demonstrated in a variety of disciplines, it has yet to be applied extensively to the field of reservoir simulation. This work shows IGA's potential as a useful tool for next generation reservoir simulation, highlighting its ability to exactly capture complex-shaped geometrical features in the reservoir and, consequentially, accurately model the pressure fields. First, an IGA code is validated against the analytic solution for a quarter five-spot pattern as well as a reference solution for a straight line source problem with pressure enforced at the corners of the domain. Next, the numerical efficiency of IGA is compared against a finite element method solution of single phase flow in a 2D representative reservoir containing a centralized "S-shaped" line source. Results suggest a clear advantage of the IGA method over the classical finite element method, which has lower convergence rates in terms of number of degrees of freedom.

This work suggests a paradigm change. Instead of history matching an (upscaled) geological model by the reservoir engineer focusing on matching the dynamic data. The model builder should be equipped with the right tools to identify areas where the geological model cannot fulfill the dynamic requirements. The target is to make history matching obsolete. Every modification of the model must have its geological reasoning and justification. Introducing artifacts for the sole target of matching dynamic data is depreciated. Often such models lack forecasting credibility. Especially selecting infill drilling locations becomes highly uncertain. This paper presents applicable and ready to use methods which will be demonstrated on a full-field project. There are two key elements, each to be fulfilled during every timestep. At first the average (region) pressures must fit and second each well must be allowed to search for the required phases to satisfy its production targets. The first requires at least one region for which the historic pressure decline is known. At the delineation towards an outer aquifer water will be automatically injected/withdrawn to match the pressure history. The second is realized by perforations (well-grid-connections) which are allowed to move away from the well's trajectory and to search for the required phases within a user or algorithm defined drainage area. There is just one restriction. Drainage areas must not overlap. In case the drainage area does not contain the necessary fluids the geological model has to be revised. At the same time important information about fluid movements are gathered. The promoted concept was successfully applied to a North African naturally fractured carbonate reservoir having more than 500 MMstb original oil in place and a 35-years production history. The reservoir was divided into eight pressure regions. Five were selected as target for the automatic pressure match, the remaining three were control regions. Two update cycles were performed. During the first cycle the structure was corrected. Thus the phase distribution was affected so that all drainage areas could produce the required historic rates. During the second cycle local adjustments in porevolume and communication across faults were performed to increase the pressure match of the control regions. The demonstrated workflow has the potential to revolutionize the commonly practiced reservoir model creation. Up to now no similar workflow allowing to assess the geological model quality within one single simulation run was published. It is unique that pressure and production targets can be met in each timestep without corrupting the underlying geological model. Contrary the modeler is provided with the required information where and how to improve the geological model.

The application of crosslinked gels has seen widespread use within the petroleum industry to divert fluid flow and remediate leakage at wells. The current study aimed at investigating the effectiveness of polymer-gel treatment in remediating a potential CO2 leakage site in the subsurface. This paper presents a brief summary of the results of laboratory characterisation work carried out on polymer gels, which involved characterisation of several polymer-gel systems with respect to their working and gelation times and conducting core flooding experiments on reservoir rocks to characterise their permeability reduction behaviour. The chemical flooding reservoir simulator, UTCHEM, was used to simulate polymer injection and its subsequent gelation process in the saline aquifer model. Parameters such as polymer concentration, polymer to crosslinker ratio and pH, and their influence on the gelation process and the area of influence have been investigated. The results have shown that lower polymer to crosslinker ratios lead to a relatively higher gel concentration. On the other hand, higher polymer to crosslinker ratios result in a relatively higher area of influence, which is mainly due to the decreased viscosity and slower gelation rate. It was also found that, as the concentration of H+ in the injection stream increases, the rate of crosslinking decreases and the lower viscosity polymer slug migrates to the far-field region of the reservoir formation.

A successful gas injection design is a challenging task in asphaltic reservoirs. Gas injection accelerates asphaltene deposition, which results in detrimental effects on field development and ultimate oil recovery. The main objective of this paper is to provide a workflow to mitigate asphaltene problems in reservoirs during gas injection. The workflow includes data gathering, fluid characterization, and dynamic asphaltene modeling using an in-house compositional reservoir simulator. The in-house simulator is capable of modeling (1) the effects of pressure, temperature, and composition variations simultaneously on asphaltene phase behavior, (2) asphaltene precipitation, flocculation, and deposition, and (3) wettability alteration due to asphaltene deposition. Three case studies were presented to investigate the effects of gas flooding, gas override, and wettability alteration on the dynamics of asphaltene precipitation, flocculation, and deposition. The simulation results showed that gas injection considerably increased the instability of asphaltene particles in the oil, and consequently, accelerated asphaltene precipitation and deposition in the reservoir. During miscible gas flooding, asphaltene mostly deposited at the front near the boundaries, where the front velocity was minimum. Moreover, asphaltene deposition occurred mainly in the bottom layer in the presence of gas override due to the lower velocity of the front in the bottom layer compared to the upper layer. Finally, the study revealed that wettability alteration due to asphaltene deposition had a major impact on the performance of the reservoir, specifically on the ultimate oil recovery, compared to the permeability reduction.

Fluid properties are often one of the key uncertainties in reservoir simulation studies. A workflow and guidelines are presented to develop an overall probabilistic forecasting framework, through which likely fluid models are validated, uncertainty ranges are estimated, and probabilistic PVT models are generated for use in stochastic black oil reservoir simulation modeling. Workflows are provided for two distinct scenarios. For scenarios with a large number of samples from the reservoir or a representative analog fluid property database, standard statistical methodologies are applied to generate probabilistic fluid models. However, in most cases where there are only limited data available, a unique approach is adopted to generate a representative pseudo-database for uncertainty assessment workflows. The pseudo-database is generated based on fluid property correlations that are screened for the conditions of the specific fluid system. The pseudo-database is used to both validate a reliable likely case and to determine ranges to generate probabilistic models. Numerical formulations of probabilistic models are provided. Application of the workflow is demonstrated for three different cases of fluid property data. For a case based on a large set of PVT measurements, standard deviation of specific fluid properties (FVF, GOR, Viscosity, and Compressibility) is calculated and used in standard statistical methods to calculate P10/P90 cases. In another example, the validity of fluid viscosity data based on an assumed viscosity correlation was investigated, suggesting that fluid viscosity was actually closer to a P20 model as opposed to a P50 (likely case). In another case study, where only limited PVT data were available, a number of screened/filtered fluid property correlations are used to represent available PVT data, and then the input parameters are perturbed to generate a pseudo-database to capture fluid property ranges as well as variation in a field. Currently, there is a lack of a systematic way to determine uncertainty ranges for fluid properties, as well as validating the most likely case (best case or mid case). The novelty of our framework is in the provision of a consistent workflow to generate probabilistic fluid models that are not based on experiential methods employed in the past.

This paper presents an optimization methodology for CO2 enhanced oil recovery in partially depleted reservoirs. A field-scale compositional reservoir flow model was developed for assessing the performance history of a CO2 flood and optimizing oil production and CO2 storage in the Farnsworth field unit (FWU), Ochiltree County, Texas. A geological framework model constructed from geophysical, geological and engineering data acquired from FWU was used for the reservoir modeling. A laboratory fluid analysis was tuned to an equation of state and subsequently used to predict the thermodynamic minimum miscible pressure (MMP). An initial history calibration of primary, secondary and tertiary recovery are conducted as the basis for the study. After a good match was realized, an optimization model with proxy was constructed with an objective function that maximized both oil recovery and CO2 storage. Experimental design was used to link uncertain parameters to the objective function. A reduced order proxy model was necessary to reduce computational cost. Control variables considered in this study included: CO2 purchase, recycled CO2, water alternating gas cycle and ratio, infill wells and bottomhole pressure of injectors and producers. Polynomial response surface methodology was used to create the proxy model based on training simulations. This involved an iterative process until a validated surrogate model was achieved. A sensitivity analysis was first conducted to ascertain which of these control variables to include in the reduced order model. A genetic algorithm using a mixed-integer capability optimization approach was employed to determine the optimum developmental strategy to maximize both oil recovery and CO2 storage. The proxy model reduced the computational cost significantly. The validation of the reduced order model ensured accuracy in the dynamic modeling results. The prediction outcome showed the robustness and reliability of the genetic algorithm in optimizing oil recovery and CO2 storage. The reservoir modeling approach used in this study showed an improved way of optimizing oil production and CO2 storage within partially depleted oil reservoirs such as FWU. This study serves as a benchmark for potential CO2–EOR projects in the Anadarko basin and/or geologically similar basins throughout the world.

Comprehensively understanding the gas hydrate accumulation mechanism is significant for the investigation of subsea gas hydrate reservoir which can provide further guidance for the hydrae exploration and development as well as the safe deep-water drilling. In this paper, a preliminary conceptual model is established to study the characteristics of gas hydrate accumulation in the typical shallow formation under the sea using a reservoir simulation method. A partial equilibrium reaction model based on the phase equilibrium of gas hydrate is used to describe the trigger mechanism of hydrate formation when methane from deep formation migrates into the upper hydrate stability zone under the seabed. Two cases are simulated for comparison, one considering the barrier effect of cap rock at seabed while the other assuming a cold spring at seabed. The simulation results indicate that in the cap rock case, a thick hydrate layer tends to be formed in the upper subsea formation but with a relatively smaller hydrate saturation, while in the case of cold spring, nearly 90% of methane from deep reservoir would leak into the sea water, nevertheless the long-term slowly gas driving water process is favorable for generating high hydrate saturation. Generally, low flux of methane gas, cap rock barrier, deep water depth, and small geothermal gradient below mud line are beneficial to forming valuable hydrate reservoirs with larger thickness and high abundance. This study has proven that the reservoir simulation method can be an effective tool to simulate the process of gas hydrate formation and accumulation in the shallow formation under the sea, which deserves for further study.

The Walloon Coal Measure (WCM) in the Surat Basin in Australia consists of coal-rich mire and a fine-grained meandering fluvial system. The main gas producing targets of WCM are numerous thin coal plies within six coal members with frequent pinching outs, splitting and merging. The geology is stratigraphically complex making correlations of individual coal plies difficult. Consequently, previous geological studies have been mostly based on coal members instead of individual coal plies resulting in inadequate description of the heterogeneity of the coal deposit. To remedy this situation, we proposed a workflow using high-resolution sequence stratigraphy to build an isochronic stratigraphy framework of sublayers and coal plies by utilizing all available data from cores and logs. The key methodology was to identify single fining-upwards cycles with coal, clay or siltstone at the top and sandstone at the base. Then similarity analysis on the cycles was used to identify aggradation, progradation or retrogradation of fluvial facies sequence between adjacent wells. Log density cutoff was used to identify coal, shaly coal, shale, sandstone and siltstone from the whole Walloon fluvial system. Reservoir parameters including gas, ash, moisture content, density, and permeability versus depth were correlated taking into consideration depth shift, regional core data and lithology in different members. All of the above were integrated into a ply-based geomodel which was used to identify highly concentrated, overlapping, continuous plies that are potential sweet pots for field development. Our intent is to provide analogue information and understanding for the coal seam distribution in the green field development of the Surat Basin. This methodology was applied to WCM to perform division and correlation of 20 sub-layers and 125 single plies with thickness ranging from 0.3–1.4 m. Coal distribution area versus thickness relationship was generated to analyze the variogram range used for some key properties, especially density and net-to-gross, and to investigate the impact of coal continuity on well spacing. Five micro-facies in fluvial system were used to describe the distribution of coal properties, impact of coal architecture and heterogeneity. Several potential sweet spots for field development were identified. With proper upscaling, this high-resolution ply-based model can be used in reservoir simulation to forecast production and calculate estimated ultimate recovery (EUR). This methodology has been applied to three coalbed methane (CBM) fields in the Surat Basin in Australia. It is novel in applying high-resolution sequence stratigraphy used in geomodel building of convention oil and gas reservoirs to CBM characterization. It has resulted in a better understanding of the complex depositional character of the WCM and consequently more accurate determination of potential sweet spots, production forecast and EUR calculation.

Originally proposed for conventional reservoirs, the dual-continuum idealization is being applied directly to simulation in ultralow-permeability reservoirs. Recent gridding techniques can mesh the geometries of large discrete natural fracture networks (NFNs) into an unstructured grid to simulate fluid flow. Insufficient literature exists to help identify the tradeoffs in selecting one approach over another. This paper not only analyzes two approaches with respect to their underlying assumptions and applicability but also proposes a hybrid approach. A commercial reservoir simulator supporting both dual-continuum and unstructured simulation grid capability was used to compare these approaches in two separate stages using CPU times and accuracy of results as metrics. First, small-scale reservoirs with simple two-dimensional (2D) fracture patterns were simulated to examine the impact of matrix permeability, fracture spacing, and fracture orientation. Second, drilling-spacing-unit-size reservoirs with networks of stochastically generated three-dimensional (3D) fracture surfaces were simulated to compare the effect of density and clustering of fractures. These models were also simulated using a hybrid approach, modeling one portion of the fracture network as discrete and the remaining as dual continuum. Results show that although the dual-continuum technique can be very fast, it is not appropriate for all simulations. For example, extremely long transient periods in ultralow-permeability reservoirs raise doubts about the applicability of steady-state transfer functions in such models, whereas this assumption is considered acceptable at higher permeabilities. Conversely, discrete-fracture techniques, where the fracture geometry is resolved accurately in structured or unstructured mesh, are applicable to a broader set of problems but can become prohibitively slow for dense NFNs. A variety of cases were explored that demonstrate the strengths and weaknesses of the two approaches. The proposed hybrid approach discussed in this paper offers a good balance between run times and the quality of results. This paper offers practical insight into technique selection over a large class of problems. Furthermore, the proposed hybrid approach of combining dual and discrete NFN simulation in the same unstructured grid is novel but provides clearly demonstrated benefits. Stochastic generation of 3D NFNs, conditioned on secondary data, is used to control the clustering of natural fractures. Two distinct types of unstructured gridding methods, both supporting multi-point flux approximation, are used to mesh discrete NFNs.

Petroleum reservoirs are geologically large and complex. In order to technically and economically optimize the exploitation of these hydrocarbon reserves, multimillion dollar investments, reliable numerical reservoir simulation models should be constructed to predict the reservoir performance and response under different production scenarios. Reservoir numerical simulation models can be only trusted after good calibration with actual historical data. The model is considered to be calibrated if it is able to reproduce the historical data of the reservoir it represents. This calibration process is called history matching, and this is the most time consuming phase in any numerical reservoir simulation study. Traditional history matching is carried out through a trial and error approach of adjusting model parameters until a satisfactory match is obtained. The biggest challenge that faces the simulation engineer during this critical phase is that, several combinations of reservoir history matching parameters might be satisfactorily matching the past dynamic behavior of the system which makes the process ill-posed due to the non-uniqueness solution issues. Traditional history matching is accordingly a time-consuming, expensive, and often frustrating procedure and as a consequence, the assisted history matching technique has been arisen. In assisted history matching technique, the simulated data is compared to the historical data by means of a misfit function, objective function. The history matching problem is translated into an optimization problem in which the misfit function is an objective function bounded by the model constraints. The objective function is minimized using appropriate optimization algorithm and thus the results are the model parameters that best approximate the fluid rates and pressure data recorded during the reservoir life. The objectives of this paper are to clarify the idea behind assisted history matching process and discuss its different aspects, experimental design, proxy models, and optimization algorithms, and show how these aspects are integrated to overcome the frustrating nature of the history matching problem. Finally, guide lines are provided to enhance the design of the assisted history matching process.

Given limited CO2 supply, operational constraints, and pattern specific reservoir performance, WAG schedule can be customized such that NPV or other metrics are optimized. Depending on the WAG schedule, recovery can fluctuate between 5–15% at the pattern scale due to reservoir heterogeneity causing variations in sweep efficiency. An analytical method was developed to optimize WAG schedules that couples traditional reservoir modeling and simulation with machine learning, enabling the discovery of optimal WAG schedules that increase recovery at the pattern level. A history-matched reservoir model of Chaparral Energy's Farnsworth Field, Ochiltree County, TX was sampled intelligently to perform predictive reservoir flow simulations and artificially build an intelligent reservoir model that samples a broad range of possible WAG scenarios for optimization. The intelligent model generates the next "best" sample to investigate in the numerical simulator and converges on the optima, quickly reducing the number of runs investigated. Results in this paper demonstrate that there can be significant improvements in net present value as well as net utilization rates of CO2 using this analytical technique. The WAG design generated by the intelligent reservoir model should be deployed in the field in early 2016 for validation. It is intended that the intelligent reservoir model will be updated on a regular basis as injection and production data is obtained. This effort represents the beginning of a paradigm shift in the application of modeling and simulation tools for significant improvements in field production operations.

Gas shales act as both the source rock and reservoir for the petroleum. One of the important characteristics of the reservoir is its low permeability, making gas production difficult. Although economic shale gas production has been achieved through horizontal drilling and multistage hydraulic fracturing, shale reservoir permeability is still one of the critical parameters in the evaluation of a shale gas play. In the present work, experimental measurement of shale anisotropic permeability is determined using a cubic shale sample in a triaxial cell. Anisotropic permeability was measured at a series of gas pressures and confining pressures. A permeability model incorporating stress and Klinkenberg effect was applied to describe the data. The model was applied in the reservoir simulator SIMED IITM to investigate the impact of anisotropic permeability and its change on shale gas production. Results are compared between using the vertical permeability, horizontal permeability, or anisotropic permeability as the reservoir permeability. The results show that using vertical permeability will significantly underestimate the gas production rate. This demonstrates that measuring directional permeability and using the most appropriate one is important for evaluating shale gas production and development of shale gas assets.

Coal Seam Gas reservoirs are naturally fractured and the flow of fluid throughout the coal occurs by diffusion through the coal matrix and then via the cleats (network of fractures) towards the wellbore. Effective permeability of the cleats and Langmuir isotherm parameters of the matrix coal blocks are the key parameters in determining the economic viability of producing from a CSG reservoir. Welltest analysis in CSG wells is more complicated than conventional reservoirs due to the stress sensitive permeability of the cleats and the great heterogeneity of coal. Fluid pressure inside the cleats initially controls the stress sensitive permeability. During production and at the reduced pressure near the wellbore region, the cleats have smaller aperture and less permeability than the cleats away from the wellbore due to the in-situ stresses. Conversely during injection, the cleats near the wellbore have higher permeability than the cleats away from the wellbore. The common well test analysis methods normally provide undefined values of average permeability and skin factor, rather than addressing the effective permeability of the cleats. This study presents how to interpret the skin factor in a CSG welltest, and also proposes a methodology using integration of welltest analysis and image log data processing results for estimation of cleat characteristics in CSG reservoirs to be input in a dual porosity CSG reservoir simulation model.

During well drilling, completion, stimulation and fracturing, moisture invasion and phase trapping lead to a drastic permeability reduction, which prevent the tight gas reservoir producing at an economical rate. To eliminate such formation damage, the power of microwave was considered. As an effective heating technique, the application of microwave heating in removing moisture in the near wellbore area is relatively novel and thus lack of experience. This paper simulates effects of microwave heating on the formation damage and gas production rate of a tight gas reservoir in Western Australia and demonstrates its effectiveness. The well information for a tight gas well in Western Australia was obtained to simulate the actual gas production in numerical model. Finite Element Analysis (FEA) software and industry standard reservoir simulation software were coupled to simulate microwave heating at borehole scale. The electrical conductivity was extracted from well logs and the dielectric constant and loss factor of the reservoir rock was calculated using the empirical equations and proven mixing law. Two cases studied are in non-fractured well and fractured well. In each study, same water amounts were injected for 5days followed by a same period of clean-up and gas production, the difference is one model has microwave heater while the other not. The effects of microwave heating on water saturation and relative permeability to gas were made between two scenarios of gas production with heating and without heating. The electric filed and magnetic field distributions and the heat generation rate in the near wellbore area were calculated based on the actual petrophysical properties of the reservoir rock and Maxwell's Equations. Based on the results from FEA, the heat generated by microwave was applied on the well models respectively. After microwave heating started, gas relative permeability increased and water saturation decreased through the heated zone. In all the cases, microwave heating seems to be an effective way to remove moisture and to recover the gas productivity back to a high rate. But in non-fractured reservoir, the gas production rate has achieved the most improvement. The present paper focuses on the interaction of microwave and trapped moisture in the reservoir. By simulating the heating assisted gas production, illustrate how microwave heating was able to increase gas well productivity by reducing water saturation. More factors need to be considered in the future, and the suggestions have been given in this paper.

The Spring Gully coal seam gas (CSG) field, operated by Origin Energy on behalf of Australia Pacific LNG, is located in the southern region of the Bowen Basin. Commercial production commenced in 2005 from the Late Permian Bandana coal seams. This paper describes the reservoir characterisation methodology employed to history match geologically unique regions of the Spring Gully field and to understand production mechanisms across the field. In addition, the paper presents a detailed discussion on how learnings from the developed region of the field has been applied to characterise the future development regions of the field for forecasting. The Spring Gully CSG field exhibits a large degree of localized heterogeneity and extensive regional geological variability. Strong gas productivity and rapid water gas ratio decline have been observed in the north-eastern region of the field. While in the western portion of the field there exists moderate gas productivity and minimal water decline. A comprehensive, multi-disciplinary evaluation of the Spring Gully CSG field reservoir and production characteristics was performed in 2014. A reservoir simulation model has been constructed based on a number of geological and reservoir characterisation studies, combined with reservoir inputs and production analysis. The integrated subsurface model forms the basis to understand Spring Gully CSG field production mechanisms and ultimately to perform production forecasts for reserve and resource estimates, for ongoing reservoir management needs and appropriately size gas and water treatment facilities. A novel approach of constructing the permeability and porosity distribution in the reservoir model was implemented. Historical well peak water and gas rate, rig testing water rate and welltest-derived permeability were used to construct the permeability distribution in the model. The permeability distribution was transformed to porosity distribution. The faulting and compartmentalisation were introduced in the north-eastern region of Spring Gully reservoir model during the history match stage to match the rapid decline of reservoir pressure. The learnings obtained from studying well production mechanisms and model history matching were subsequently applied to the undeveloped regions of the field. Extensive validation and quality control of forecast methodology, forecast parameters assumptions and life-of-field production forecasting were performed.

This paper demonstrates the multivariate approach as a tool for the management of initial data uncertainties in 3D reservoir simulation processes. The multivariate approach helps to address several problems, including optimizing the development scenario, determining the well type, and calculating the optimal horizontal well completion system. The method for determining the optimal development scenarios includes performing multiple forecast runs on simulation models with variations of initial data (such as permeability and water/oil contact) for various development scenarios (including well spacing and development plan). Then, the data are systematized, and a statistical analysis is performed with the calculation of economic indicators. The distributions of technical and economic development parameters are built from the results of these processes. The optimal development scenario is selected directly by selecting the realization with the maximum probability of occurrence from the most economically successful realizations. In other words, a development scenario is considered optimal if it has maximum probability of occurrence in the data set and maximum economic indicators (net present value (NPV) and profitability index (PI)). The paper uses two case studies to illustrate the multivariate approach for determining the optimal development scenarios by considering the initial data uncertainties. For gas field A (eastern Caspian, lower Jurassic terrigenous deposits), the task of determining the optimal well type and spacing of the new wells was addressed by using multivariate reservoir simulation. The data uncertainty is attributable to a core test result error in the permeability determination and to a lack of data for the gas/water contact depth in accordance with the well testing and production testing results. A total of 10,000 simulations of the multivariate reservoir and economic model were run to determine the optimal well type and well spacing. In addition, probabilistic values of the development technical and economic indicators P10, P50, and P90 were determined. The results of the calculation enabled the optimal development parameters to be selected and the effect of the initial data uncertainties on the final technical and economic indicators of the field development to be evaluated. The length of horizontal wells, hydraulic fracture spacing, and optimal well spacing were selected based on the multivariate approach for oilfield B (western Siberian oil and gas province, upper Jurassic). For this purpose, 100,000 iterations of the development scenarios forecast were calculated based on the multivariate simulation model. Based on the analysis of the calculations and economic model, the optimal length of the horizontal section and fracture spacing were selected and justified, enabling the optimal design and completion system for horizontal wells, as well as risk mitigation measures, to be selected.

The in-situ CO2 huff-n-puff (ISCHP) is a promising EOR technique that could overcome some major drawbacks associated with continuous CO2 flooding, e.g., early CO2 breakthrough, high operation costs, demand for CO2 source, etc. Field pilot test of in-situ CO2 huff-n-puff was carried out in Jiangsu Oilfield based on lab experimental study, which results in the augmentation of well productivity and water cut reduction. Our work is to develop the kinetics model of gas generation reaction of the pilot and discuss ISCHP mechanisms base on reservoir simulation. Through in-laboratory experiments, a salt solution with surfactant additives (gas generation reagent) was injected into the sand pack model to evaluate ISCHP efficiency. The kinetics model of gas generation reaction was developed by matching the decomposition experiment results of the pilot. According to the reservoir geological characteristics, a reservoir model was built and calibrated based on the past production observations of the pilot. Furthermore, a series of sensitivity analysis of the operating parameters, including concentration of the reagent, injection volume and soaking period, were conducted to obtain the optimum scheme. Reservoir simulation illustrates the reagents can generate large amount of CO2 near the wellbore formation. ISCHP process can increase oil production by 56%, and decrease the reservoir temperature near wellbore by 5°C with the endothermic decomposition. The most influential parameters are concentration of the reagent and injection volume. Besides, the optimum scheme is achieved with a 16.4 wt% concentration of the reagent, an injection volume of 250t, and a soaking period of 7 days. This study can provide fundamental understanding of ISCHP enhanced oil recovery in the Jiangsu oilfield.

Gas Flooding is one of the reservoir production management methods which commonly used in the petroleum industry in order to improve oil recoveries and maintaining reservoir pressure. This study discusses two methods in analyzing gas flooding, i.e., material balance analysis and reservoir simulation methods. The objective is to examine the ability and advantages of material balance method with respect to reservoir simulation. The studied field has three productive zones which consist of two light oil zones and one dry gas zone. Gas flooding scenario was conducted in one light oil reservoir and was modeled using both material balance and reservoir simulation methods to predict reservoir performance for about 14 years and the results were compared to investigate the effect of gas injection. The material balance method employs production, PVT, pressure data, and averaged properties of geological model for a single tank model. On the other hand, the reservoir simulation method uses a compositional model to investigate in detail the effect of injected gas to reservoir fluid behavior and the oil recovery The examinations indicate that the material balance analysis method shows the advantages of a requiring much less data, but could give reasonably accurate results. Furthermore, the application of material balance analysis in gas flooding could be used for gas injection rate optimization for a very short time manner and could assist for further development of reservoir simulation. For future development, material balance analysis could be used to analyze gas flooding scenario to save time and cost rather than reservoir simulation which takes longer time.

This paper presents our work on developing a platform for high performance reservoir simulations, which is developed to support the implementation of various reservoir simulators on distributed-memory parallel systems. This platform employs MPI (Message Passing Interface) for communications and OpenMP for shared-memory computation. It provides structured grids due to its simplicity and cell-centered data for each grid cell. The platform has a distributed matrix and vector module and a map module. The map connects the grid and linear system modules. Commonly-used Krylov subspace linear solvers are implemented, including the restarted GMRES method and the BiCGSTAB method. It also has an interface to a parallel algebraic multigrid solver, BoomerAMG from HYPRE. Parallel general-purpose preconditioners and special preconditioners for reservoir simulations are also developed. The numerical experiments show that our platform has excellent scalability and it can simulate giant models with hundreds of millions of grid cells using thousands of CPU cores.

The classical discontinuous Galerkin (DG) methods are designed for elliptic (parabolic) problems and hyperbolic problems. For reservoir simulations, the pressure equation from the black oil model is elliptic (parabolic), while the equations for saturations are hyperbolic. Due to this special property, it is difficult to directly apply the discontinuous Galerkin methods to the black oil model. In this paper, we extend the discontinuous Galerkin methods to reservoir simulations. In our schemes, the local discontinuous Galerkin (LDG) method is used to discretize the black oil model. The upwind concept is combined with the numerical flux term of the LDG method to simulate the direction of propagation of the multiphase flow in reservoirs to avoid the unphysical solutions. We also extend the Peaceman model to the discontinuous Galerkin methods on unstructured grids. Based on the extended discontinuous Galerkin methods, we employ the iterative implicit pressure-explicit saturation (iterative-IMPES) and fully implicit (FIM) methods to solve the coupled nonlinear black oil model. A parallel simulator is implemented using the parallel adaptive finite element package, Parallel Hierarchical Grid (PHG), and validated by testing the first and ninth SPE Comparative Solution Projects. The parallel scalability of our simulator is also tested by a large scale case.

Cold heavy oil production (CHOP) exploits the mechanism of enhanced solution gas (or foamy-oil) drive to achieve an economic oil rate and ultimate recovery. Understanding the foamy-oil dynamics and being able to simulate the process enable us to evaluate the cold production potential more realistically. This has led to the improved cold heavy oil recovery in Shell Canada's Cliffdale development in Peace River area. A recently developed foamy-oil dynamic model has been employed to evaluate the effects of well spacing and patterns on oil recovery. The results revealed that under the foamy-oil drive, the oil recovery improves as the well spacing decreases, because of decreasing well spacing leads to faster reservoir pressure depletion and stronger foamy-oil drive. In contrast, with a conventional black-oil model, the estimated ultimate oil recovery stays constant irrespective of the well spacing, and the only benefits of down spacing would be the production acceleration. The benefits of capturing foamy oil dynamics for the evaluation of CHOP development has been demonstrated with an example of a high level economic screening approach for the search of optimal well spacing and the number of laterals. The evolution of typical well spacing and the number of well laterals with time in the Peace River CHOP development has resulted from both the ever-increasing field operation experience and the improved understanding of foamy-oil drive dynamics.

Managing uncertainties during subsurface modelling in brown field re-development requires robust identification and quantification of impact of the underlying uncertainties. Within a given cycle of integrated reservoir modelling, a modelling strategy can only be defined based on associated uncertainties and development options. This paper focuses on using the Design of Experiment to screen and quantify the impact of uncertainties and evaluate development outcomes in a brown field. The paper details steps taken to identify and quantify subsurface uncertainties in a multi stacked reservoirs that could impact development options. The overall development strategy was to: (a) Introduce artificial lift and offtake management to keep existing wells flowing, (b) Install gas lift supply lines from new gas-lift skid at the flowstation to existing wells requiring gas-lift, (c) drill and complete additional wells using existing locations, (d) hook up the new wells to existing remote manifold and route to flowstations via existing bulklines. To access the impact of uncertain parameters and its ranges of uncertainty were identified and quantified based on current understanding of the reservoirs. The Plackett-Burman Design of Experiment was used to screen each parameter using the Tornado and/or Pareto Plots. The key uncertainties (heavy hitters) identified from the screening stage were carried forward to develop a Response Surface Model (RSM) using Box-Behnken experimental design in order to sample the full uncertainty space associated with each reservoir. The probability distributions of In-Place and cumulative production were generated using Monte-Carlo analysis and estimate of the Proved, Probable and Possible volumes and ultimate recovery were obtained (part of methodology) From the study results, the feasibility for further oil and gas development based on 3D reservoir simulation with several development scenarios and options were evaluated. Results of the deterministic possible outcomes were used to identify specific cases that closely matched the Proved, Probable and Possible volumes from the Monte-Carlo distribution. Model provides tool for better well and reservoir management.

Cold heavy oil production exploits the mechanism of enhanced solution gas (or foamy-oil) drive to achieve an economic oil flow rate and ultimate recovery. Understanding the dynamics of foamy-oil drive and being able to simulate the process make it possible to evaluate the potential of cold production more realistically. A generic kinetic model based on known solution gas exsolution and evolution process has been proposed and tested. The sequence of gas exsolution in foamy-oil drive can be described by the following four major steps and corresponding (first-order) mechanisms: Supersaturation from pressure drawdownBubble nucleation under super-saturationBubble growth under molecular diffusionBubble coalescence from film drainage The effects of dispersed gas on the gas phase mobility has been modeled with a simple viscosity mixing rule by assigning a much higher viscosity to the dispersed gas component. The benefits of this approach is that the apparent low mobility of the gas phase in the foamy-oil drive process can be attributed to the viscosity change, instead of rate dependent relative permeability. Therefore, a “normal” relative permeability can be used for a wide range of rate conditions. The proposed dynamic model has been validated with the history matches of both the volume draw down and the pressure draw down laboratory test data. The validation reveals that one can use a set of consistent dynamic parameters to match the results of various tests under different draw down rates. The dynamic model has been implemented in a commercial reservoir simulator (CMG-STARS) with available options.

Nowadays steam injection is commonly used as a thermal EOR method due to its efficiency for recovering hydrocarbons from heavy oils and bitumen. Reservoir models simulating this process describe the thermal effect of the steam injection, but generally neglect the chemical reactions induced by the steam injection. These reactions are called aquathermolysis and have been previously investigated through laboratory experiments. Based on these experimental results, this paper proposes a new compositional kinetic model to reproduce these reactions in the context of reservoir modeling. In particular the proposed reactions model accounts for the formation of the highly toxic and corrosive acid gas H2S in the presence of sulfur-rich heavy oil and predicts the modification of the oil SARA composition versus time that results from the aquathermolysis reactions. The overall objective of this paper is to understand the aquathermolysis reactions in reservoir undergoing steam injections and to provide the Oil companies with a numerical model for reservoir simulators that forecasts H2S production risks. The new Sulfur-Based-Compositional Kinetic Model is firstly presented and then validated in the context of laboratory-scale experiments. Its H2S and SARA composition predictions are based on a new simplified reactive scheme of 5 reactions coupled to a compositional and thermal reservoir simulation with no mass transfer (0D). Finally the thermo-kinetic modeling is coupled with a 2D reservoir simulation of a SAGD process for a generic Athabasca oil sand. The H2S to SARA (bitumen) production ratio against time computed by the reservoir simulation is found to be consistent with data from a SAGD process in Athabasca, which validates the modeling approach followed.

Chemical enhanced oil recovery (CEOR) of heavy oils is growing in volume and scope due to advances in the technology and field experience. This work describes a new methodology to select a CEOR strategy in a heavy oil reservoir when several viable options exist. We applied this methodology to the Pelican Lake field in Alberta. We evaluated water flooding, polymer flooding, alkaline-surfactant-polymer (ASP) flooding, alkaline-co-solvent-polymer (ACP) flooding and polymer flooding followed by ASP flooding in laboratory tests. We executed new experiments including microemulsion phase behavior, polymer rheology and corefloods representing these various strategies. These experiments were designed to help understand the role of mobility control in chemical flooding of heavy oils. UTCHEM, the University of Texas Chemical Flooding Simulator, was used to model experimental results, and to scale them up in pilot simulations using heterogeneous geological models representative of Pelican Lake. We report results for the selection of promising CEOR strategies for implementation in Pelican Lake based on the new laboratory experiments, reservoir simulations and our qualitative understanding of their various advantages and disadvantages. We present simulation results of a pilot using horizontal wells in a heterogeneous geological model representative of the reservoir. We simulated the various chemical EOR processes using the matched experimental data and evaluated them in terms of total oil production, time to completion and complexity. In-situ oil viscosity and operational injection limits were evaluated as crucial sensitivities. We make recommendations for CEOR implementation based on simulation study results and our understanding of relative process risks and costs.

The accuracy of Equation of State (EoS) in phase behaviour modeling plays an important role in compositional and thermal reservoir simulations. PC-SAFT (Perturbed-Chain Statistical Associating Fluid Theory) EoS can predict solubility and density of bitumen/solvent systems more accurately than Peng-Robinson and Soave-Redlich-Kwong EoS. In this work, a simplified version of PC-SAFT EoS was integrated into a thermal reservoir simulation to model heat assisted solvent recovery process. Bitumen was characterized into two scenarios for the simplified PC-SAFT modeling: eight pseudo-components (8-PCs) and three pseudo-components (3-PCs) using a well-developed bitumen characterization method. The solubility and density modeling ability of 3-PCs was evaluated and verified with that of the 8-PCs system. The simplified PC-SAFT with 3-PCs bitumen characterization method was then applied to generate data of K-Value, density, compressibility, and thermal expansion coefficient of each pseudo-component. The generated fluid properties were integrated with reservoir simulation to model a warm C3H8 VAPEX experiments at 40, 50 and 60 °C. In the 8-PCs characterization, the overall Average Absolute Relative Deviations (AARDs) for the solubility and density of CH4, C2H6, C3H8 and CO2-saturated bitumen were within 6.6 % and 2.3 % respectively, and the predicted AARDs for the density of n-C5H12, n-C10H22, n-C14H30, toluene and xylene-diluted bitumen were as low as 0.9 %. Results of the 3-PCs characterization were found to be as accurate as those of 8-PCs in modeling bitumen/gas solvent systems. In addition, the results of the fluid properties were successfully integrated into commercial reservoir simulator. Relative permeability was history matched with the measured data at 40 °C, and then tested against the measured data at 50 and 60 °C. The reservoir simulation results indicated good agreement with the measured data. The numerical results showed that although bitumen viscosity decreased due to high temperature solvent, solubility was reduced at the same time, compensating the viscosity reduction effects by solvent diluting.

Typical thermal processes involve sophisticated wellbore configurations, complex fluid flow and heat transfer in tubing, annulus, wellbore completion, and surrounding formation. Despite notable advancements made in wellbore modeling, accurate heat loss modeling is still a challenge using the existing wellbore simulators. This challenge becomes even greater when complex but common wellbore configurations such as multi-parallel or multi-concentric tubings are used in thermal processes such as Steam Assisted Gravity Drainage (SAGD). To improve heat loss estimation, a standalone fully-implicit thermal wellbore simulator is developed that can handle several different wellbore configurations and completions. This simulator uses a fully implicit method to model heat loss from tubing walls to the surrounding formation. Instead of implementing the common Ramey method (1962) for heat loss calculations that has been shown to be a source of large errors, a series of computational fluid dynamical (CFD) models are run for the buoyancy driven flow for different annulus sizes and lengths and numbers of tubings. Based on these CFD models, correlations are derived that can conveniently be used for the more accurate heat loss estimation from the wellbore to the surrounding formation for SAGD injection wells with single or multiple tubing strings. These correlations are embedded in the developed wellbore simulator and results are compared with other heat loss modeling methods to demonstrate its improvements. A series of validations against commercial simulators and field data are presented in this paper.

In waterflooded reservoirs under active scale management produced water samples are routinely collected and analysed, yielding information on the evolving variations in chemical composition. These produced water chemical compositional data contain clues as to the fluid/fluid and fluid/rock interactions occurring in the subsurface, and are used to inform scale management programmes designed to minimise damage and enable improved recovery. In this interdisciplinary paper, the analyses of produced water compositional data from the Miller Field are presented and a 1D reactive transport model is developed to study possible geochemical reactions taking place within the reservoir through matching model results with observed produced water data. However, in the 1D reactive transport model, only one flow path was simulated; this does not fully represent the fluid flow and mixing behaviour in the reservoir. Therefore, this paper also presents a fully 3D reservoir simulation study for the Miller Field to evaluate brine flow and mixing processes occurring in the reservoir, using an available history matched streamline reservoir simulation model integrated with produced water chemical data. Conservative natural tracers were added into the modelled injection water, and then the displacement of injection water and the behaviours of the produced water in two given production wells were further studied. In addition, the connectivity between producers and injectors was investigated based on the comparison of production behaviour calculated by the reservoir model with produced water chemical data, and an assessment of the properties of the intervening faults was also performed. Finally, a model of BaSO4 scale precipitation was included in the model, and the simulation results with and without barite precipitation were compared with produced water chemical data (observed barium and sulphate concentrations in the produced brine). In general, the modelled and observed data were found to be in good agreement, but any discrepancies were in fact found to be very informative also. The model assumes scale deposition is possible everywhere in the formation, whereas in reality the near production well zones were generally protected by scale inhibitor squeeze treatments, and thus the discrepancies between modelled and observed data could be used to diagnose the effectiveness of the chemical treatments to prevent formation damage around the production wells.