Dispersion of a solvent into heavy oil and bitumen in porous media has special significance in the context of solvent-based as well as solvent-aided versions of SAGD and CSS recovery processes. While solvent is injected with steam, the mixture condensation temperature changes based on solvent partial pressure. In addition, water condensate creates a film which acts as a barrier and impacts solvent dissolution in oil. The solvent, which is not soluble or has very low solubility in water, may not be able to diffuse in the oleic phase due to the presence of the water film. The objective of the present study is to investigate the pore-scale solvent diffusion in oil for solvent-based and hybrid (steam + solvent) processes through the following steps: Developing a pore-scale simulator, capable of handling steam and solvent condensation as well as mass transfer in porous media. Investigating solvent dissolution in the oleic phase at the pore-level considering the asynchronous condensation of solvent and steam. Investigating the dissolution of solvent, either in gaseous phase or in the form of liquified thin bulk films of condensed solvent and water condensate, in the oleic phase. A pore-scale simulator was developed with the capability of modelling solvent mass transfer and condensation of both solvent and steam, along with a Navier-Stokes type solution for the velocity field. In addition, conjugate heat transfer was included in the model that takes into account the heat transfer from solvent and steam to the solid grains by considering the two media (i.e. solid and fluid) for the solution. A realistic description of a 2-dimensional porous medium is used for direct numerical simulation (DNS). The properties of a typical heavy oil and solvent were implemented in the model with diffusion coefficient as functions of both temperature and solvent concentration. After model setup, the newly added features of mass transfer and conjugate heat transfer were validated by comparison with analytical models. For mass transfer validation, the numerical results were in agreement with analytical solution for a capillary whereas the model performance for conjugate heat transfer were inline with the analytical solution proposed for heat flow over a slab. The pore-scale simulator was then used to model two-dimensional pore-scale experiments of solvent co-injection with steam. To reproduce the experimental results, the interface advancement velocity was calculated as an evidence of the chamber growth. The 2D numerical simulation results were in agreement with the experimental data. The condensation of solvent vapor and steam also changes fluid flow and flow pathways of solvent at the pore-scale which results in some complex fluid flow and behavior such as excessive unexpected channeling. The present study is the first of its kind which considers condensation of steam and solvent vapor at the pore scale. The model is used to investigate solvent vapor condensation in competition with steam at the pore-scale and to study the impact of solvent type and operating conditions such as pressure. The outcome of the present study improves our understanding of mass transfer in porous media for solvent- based and solvent-aided thermal recovery processes.

This paper proposes a new criterion to estimate the impact of lean zones on thermal or thermal-solvent processes, such as SAGD, SA-SAGD, or EBRT (Imperial Oil's "Enhanced Bitumen Recovery Technology"). The interface stability of the steam/solvent chamber in the lean zone is related to the competition of heat transfer in either conductive or convective forms. This competition is described by a non-dimensional number, the Rayleigh number in porous media, which can be calculated by the parameters of the lean zone thickness, water mobility, thermal diffusivity, and steam chamber temperature. When a system's Rayleigh number is larger than a critical value, the primary heat transfer mechanism in the lean zone will be convection, in which the interface becomes unstable and may cause significant loss of steam or solvent. Analytical solution to obtain the critical Rayleigh number was given in literature ( Philip, 1982 ). For long axis horizontal cross sections, the critical Rayleigh number is estimated between 0.1 and 1. A significant number of simulation cases have been analyzed in a sensitivity study with CMOST (Computer assisted Matching, Optimization, and Sensitivity analysis Tool, Computer Modeling Group) to validate the above hypothesis. Monte Carlo simulation results have shown a strong correlation between the solvent recovery and the product of water mobility and lean zone thickness. This product is the essential part of the formula of modified Rayleigh number. Further calculation indicated that the critical Rayleigh number for the simulation models is between 0.1 ∼ 1 for SA-SAGD or SAGD and about 0.17 ∼ 0.32 for EBRT. Both are within the range given by the analytical solution. The critical Rayleigh number is introduced to the lean zone analysis for the first time. This physics-based criterion is able to give an answer to the estimate of performance impact of the lean zone for thermal and thermal-solvent gravity drainage processes. The Monte Carlo simulation confirmed the consistency between the numerical studies and the analytical solution.

Gas-lift is an important artificial lift strategy for increasing the production of hydrocarbons from heavy oil and offshore reservoirs with declining pressure. The optimum design and operation of gas-lift has a considerable impact on the optimum production and economics of the entire field and can only be achieved by considering all related variables in connected reservoirs, gas-lifted wells and facilities. Therefore, it is essential to formulate the gas-lift model as an optimization problem within an integrated modeling environment, where all the time dependent physical and operational constraints of reservoirs, wells and facilities can be collectively taken into account during the time evolutionary modeling of the asset. However, this could pose a computationally challenging problem for most derivative based optimizers, as some of the governing equations and models representing the gas-lifted asset could be very nonlinear, physically or mathematically discontinuo us, and the system may not have a solution under some of the conditions proposed by the optimizer due to physical operational restrictions or infeasibility. To overcome these problems we applied a mesh adaptive direct search (MADS) derivative free optimization technique to optimize the gas-lift problems in an integrated reservoir, wells and production facilities environment. The results proved the suitability of the MADS strategy for optimizing these systems particularly for systems with narrow or discontinuous boundaries of physical operation.

The conventional approach used to account for geomechanical effects when modeling steam injection operations is based on dilation/compaction tables. These tables are mostly empirical and are usually employed as a history-matching parameter to calibrate simulation models to observed injection/production rates and pressures. In this study, we use a more rigorous workflow that couples a geomechanical simulator to more accurately model the changes in permeability and pore volume that occur during high-pressure steam injection. This paper demonstrates a new approach for incorporating and modeling the geomechanical effects observed while injecting steam for heavy oil recovery, specifically during SAGD (Steam-Assisted Gravity Drainage) operations. This workflow does not need an explicit, empirical table relating pressure to permeability and porosity; instead it uses a mechanical simulator to determine the evolution of the reservoir's stresses and strains to calculate a new property distribution that is updated in the reservoir simulator to account for dilation and compaction phenomena. To model the complex thermo-poro-mechanical coupling that prevails during thermal stimulation processes, we use a numerical workflow that integrates finite-difference reservoir and finite-element geomechanical simulators. This coupling technique enables a more rigorous modeling of the fluid and heat flows while predicting their influence on the reservoir deformation and stresses. Consequently, a direct link between stress/strain and porosity/permeability can be used to model the geomechanical changes that occur in the reservoir during high-pressure steam injection. We also compare the predicted behavior of SAGD models using the coupled approach against the use of empirical dilation/compaction pressure tables. A comparison of the simulation results obtained using the proposed coupled approach versus those obtained using empirical tables (uncoupled) showed significant differences in steam injectivity and distribution as well as oil and water production.

For methods of thermal heavy oil recovery, an alternative approach of applying electrical energy, such as electromagnetic heating, can be used to generate heat in reservoirs that are not suitable for steam injection or to improve the economics and reduce environmental impact of the heavy oil recovery compared to using steam injection. While much progress in the development of electromagnetic heating technology has been made in recent years, the ability to accurately and effectively mathematically model the application of an electromagnetic heating process to a reservoir has been limited. In this paper, based on our reliable and efficient electromagnetic heating simulator, the effects of operational parameters on electromagnetic heating performance are investigated including evaluation of antenna location, well constraints, and applied power and frequency. The feasibility of electromagnetic heating in oil sands reservoirs has been examined for two cases: a) a single horizontal well containing a heating source and b) a horizontal well-pair with heating sources located in both wells.

Oil sand reservoirs play an important role in the economy of Canada due to their significant recoverable reserves. Due to the high viscosity of the oil in these reservoirs, conventional methods cannot be used for production. The steam-assisted gravity drainage (SAGD) method is an efficient way of producing oil from these reservoirs. Predicting oil production and steam injection rates is required for planning and managing a SAGD operation. This can be done by simulating the fluid flow with flow simulation codes, but this is very time consuming. The run time for a 3D heterogeneous model with one well pair can exceed 2 days. In this paper, a SAGD approximate simulator for predicting SAGD performance with 3D heterogeneous models of geologic properties is developed. This approximate simulator can handle different types of operating strategies. The approach is an approximate solution using a semi analytical model based on relevant theories including Butler's SAGD theory. The approximate simulator or proxy is much faster than the full simulator and it gives accurate estimated oil production and steam injection rates at different time steps. Theoretical and numerical research has been undertaken to develop the proxy, implement it in fast code, demonstrate the accuracy of prediction and apply to realistic examples. This proxy is used for other applications such as transferring uncertainty for reservoir realizations and well trajectory optimization.

In-Situ Reflex (ISR) is a novel solvent-based process that utilizes resistive electric heaters to vaporize solvent and recycle mobilized water downhole. ISR promises a significant reduction in greenhouse gases emissions through the elimination of steam generation and water handling facilities at the surface as well as effectively vaporizes the injected fluid along a wellbore. However, the economic viability of this process is highly dependent on the in-situ refluxing of the solvent which requires an in-depth understanding of the process and associated challenges numerically and analytically. In modeling the SAGD process, optimal operating conditions rely on a relatively constant temperature profile across the major portion of a steam chamber that leads to an excessive energy input requirement. However, ISR optimal operating conditions tend to exhibit different temperature profiles as a result of changing thermal recovery to a solvent diluting mechanism. As such, employing a SAGD analytical model results in misunderstanding the ISR fundamental thermodynamics and hindering further optimization of the process. This paper is the first time that an unsteady-state semi-analytical model has been developed for predicting ISR performance and shared publicly. The developed model has been validated using numerical simulation data and is capable of properly predicting a temperature distribution in a steam-solvent gaseous chamber in the presence of a fixed source of heat in an injector. This model includes fixed heat sources in both injectors and producers to represent the resistive heater concept, capture the reflux concept, and evaluate the contribution of refluxed solvent to reducing the solvent usage. In addition, the model helps better understand the phase behavior and the effectiveness of various solvents in further analyzing and determining the optimum downhole operating conditions and improving the overall ISR performance and its economic viability. The proposed model brings an insight into analytical modeling of the ISR process with the aim of increasing an understanding of the heat transfer mechanism, along with identifying the advantages and limitations of using the bottom-hole resistive heater technology. This will lead to a higher predictability of successful field implementation, lower upfront capital cost, higher energy efficiency, and environmentally sustainable development.

In this paper, techniques have been developed to determine multiphase boundaries of solvent(s)/heavy oil mixtures at high pressures and elevated temperatures in pressure-temperature ( P - T ), enthalpy-temperature ( H - T ), and pressure-enthalpy ( P-H ) phase diagrams, respectively. Theoretically, the Peng-Robinson equation of state (PR EOS) incorporated with a new alpha function as well as the previously developed enthalpy calculation algorithm are used to predict the multiphase boundaries of the solvent(s)/heavy oil systems by characterizing the heavy oil as a single pseudocomponent. The PR EOS associated with the newly developed alpha function is found to be accurate in reproducing the experimentally measured VL 1 L 2 (V represents the vapour phase, L 1 denotes the high density hydrocarbon-rich liquid phase, and L 2 refers to the low density CO 2 liquid phase) three-phase boundary pressures with an overall absolute average relative deviation (AARD) of 2.01% and maximum average relative deviation (MARD) of 3.54%, respectively. In addition, the VL 1 L 2 three-phase boundaries are expanded and tended to move toward the region with higher temperatures and lower pressures in the P - T phase diagram with the addition of either C 3 H 8 or n -C 4 H 10 to CO 2 /heavy oil systems compared to those of only CO 2 exists in heavy oil systems. The enthalpy changes rapidly with variation of temperature in both H - T and P-H phase diagrams within the three-phase region. Either two-phase (VL 1 ) or three-phase (VL 1 L 2 ) isenthalpic flash calculations can be performed straightly to determine phase fractions and compositions together with system temperature without conducting the stability analysis for a provided solvent(s)/heavy oil mixture at a constant enthalpy and pressure with the constructed H - T phase diagram.

Reservoir simulation projects are ubiquitous across the oil and gas industry. A less common practice is retrospective validation of the accuracy and robustness of the original model's predictions. This paper highlights the importance of validating previous simulation work with field data in order to test predictability and acquire more confidence in future simulation work. A fully implicit coupled wellbore-reservoir simulator was used to history match the performance of four steam-assisted gravity drainage (SAGD) well pairs after one year of operation. The purpose of this effort was to build a representative model that mimicked the observed field behavior and captured the key performance drivers such as reservoir quality, completion type, and operating strategy. A representative history matched model was achieved with an overall accuracy within 9.5% of actuals after one year of operation. This model was used to forecast well pair performance after three years. After three years of field operation, the predicted simulation forecast was compared against the actual field data. The accuracy and robustness of model's predictions remained valid and improved to within 7.5% of actuals after three years of operation. For two out of the four well pairs, the observed trends between the actuals and the simulation appeared to deviate due to completion type and operating strategy changes that occurred after the original history match was completed. Once these changes were reflected in the simulation model, predictability was restored and improved reaching an accuracy within 4.7%. This paper aims to deliver two key learnings and best practices. First, a coupled wellbore-reservoir simulation approach was successfully applied to build a representative model that captures the key performance drivers and observed field behavior. Second, a retrospective validation of the model's predictions was completed and resulted in more accurate, robust, and confident performance predictions. These robust models can be used to better understand, forecast and optimize the productive potential of oil and gas operations over their life cycle.

Warm solvent injection (WSI) has been proposed as a promising alternative to steam-based methods for bitumen recovery, due to its potential to reduce greenhouse gas emissions and environmental footprint. It involves injecting heated vaporized solvent (low molecular weight hydrocarbons) to reduce the viscosity of bitumen via solvent diffusion and latent heat transfer. However, similar to its thermal counterparts, the WSI recovery response is also highly sensitive to the underlying reservoir heterogeneities, e.g., shale barriers; in particular, the conformance of solvent chamber advancement can be a severe concern in heterogeneous reservoirs. Therefore, it is essential to approximate and monitor the development of solvent chamber during production and to optimize the operations design. Conventional monitoring methods, such as 4D seismic, can be quite costly. This work proposes a novel approach involving machine-learning techniques to efficiently track the solvent chamber positions in heterogeneous reservoirs. First, a detailed sensitivity analysis is performed to examine the impacts of shale barriers on WSI production responses, which include the oil rate and the evolution of solvent chamber. A set of synthetic simulation models for the WSI process are constructed. Petrophysical, fluid and operational variables representing typical Athabasca oil sands conditions are assigned. Different configurations of shale barriers with varying sizes, correlation length and proportions are assessed. Next, a large training dataset consisting of many heterogeneous models and their simulation results are assembled. The inputs features are extracted from the oil production based on several time-series analysis methods; the output parameters are formulated to represent the dynamic evolution of solvent chamber. Different dimension reduction and parameterization strategies are formulated and tested to represent the solvent chamber locations and interfaces. Convolutional neural network is implemented to dynamically track the solvent chamber positions by correlating the extracted inputs and outputs. The simulation results confirm that the presence of shale heterogeneities would impede the development of solvent chamber, causing a reduction in oil rate. In particular, the shale barriers that are located closer to the well pairs would exert a more severe impact on production responses than those that are located at further distances from the wells. The application of machine-learning algorithms enables the locations of the solvent chamber as a function of producing time to be inferred and tracked reliably. The proposed workflow provides a practical workflow to estimate the real-time solvent chamber development corresponding to the WSI process in heterogeneous reservoirs from oil production and solvent profile directly. The presented workflow offers a novel alternative to infer the development of solvent chamber in heterogeneous reservoirs from production time-series data directly. This type of analysis could complement many existing monitoring techniques to deliver a more comprehensive inference of the distribution of shale heterogeneities in solvent-based bitumen recovery operations. Production data is used directly to assess the conformance of solvent chamber advancement, which is an important consideration in operations design and real-time optimization.

Water-hot gas injection that uses the waste heat from existing thermal operations can be substantially more energy efficient than CSS and SAGD which require steam generation and water treatment. This project investigated the potential of water-hot gas injection to enhance heavy oil production from Lloydminster heavy oil reservoirs and reduce GHG emissions. The co-injection of water and hot gas at higher temperatures provides a drive component, increases the reservoir temperature, and leads to substantially improved oil recovery as compared to primary production. Reservoir simulations (using CMG STARS) were performed for a Lloydminster greenfield heavy oil reservoir with 3 horizontal wells. Water-hot gas injection following primary production and 3 CSS cycles was evaluated for pressure drive applications. Flue gas temperatures and compositions in the field depend on the combustion fuel and O 2 concentration used for combustion. Different mole% water values (0 to 90%), N 2 /CO 2 ratios (0 to 7.43), and temperatures (80 to 450 °C) were investigated in order to evaluate many of the available options. Although oil production for water-hot gas injection was much lower than for steam-only injection, the energy injected/m 3 of oil produced was reduced by an order of magnitude and this energy was recovered from waste flue gas. For water-hot gas injection, both oil production and energy injected/m 3 of oil increased with increasing injection temperature, increasing mole% water, and decreasing N 2 /CO 2 ratio in the injection fluid. The steam oil ratio (SOR) was ~0.6 m 3 CWE/m 3 oil and decreased gradually with injection temperature. Water-hot gas was more effective than hot gas injection as water addition to flue gas increased oil production due to the heat of vaporization that it transferred to the reservoir. CO 2 in the hot gas was more effective than N 2 . Decreasing the N 2 /CO 2 ratio increased oil production because of the greater solubility of CO 2 and its higher molar specific heat as compared to N 2 . The difference in oil production between water-hot gas and hot gas injection increased with temperature. There was an optimum hot gas injection rate as too high a rate interfered with oil flow and reduced its production. Use of vent wells to remove excess gas and injector well pulsing were unsuccessful operating strategies. The potential of using waste flue gas energy for water-hot gas injection in Lloydminster greenfield heavy oil reservoirs to enhance oil production as compared to continued primary production, improve energy efficiency, and reduce GHG emissions was demonstrated. Reservoir simulations were also performed to determine the effect of water-gas injection following CHOPS. Wormholes, which grew extensively from vertical wells during primary production, provided enhanced reservoir access, substantially improved injectivity, and supported transfer of heat into the reservoir. Both pressure drive and cyclic water-hot gas injection processes significantly increased oil production as compared to continued primary production following CHOPS. The pressure drive water-hot gas process was substantially better than cyclic injection as a result of the formation of an oil bank and downward displacement of oil by hot gas. The recovery factor for water-hot gas injection in drive mode was as high as 25.8% for 20 mole% water and 80 mole% gas at 185 °C and 30.5% for 80 mole% water and 20 mole% gas at 160 °C. Heavy oil production increased with increasing injection temperature.

Over the last decade, steam assisted gravity drainage (SAGD) process has been successfully commercialized in Alberta and Saskatchewan thermal projects. Remediation of slotted liner wells in cases of erosion, or high gas/steam production later in life requires a different basis of design based on specific well challenges and will be challenged by the additional pressure drops due to poor or variable sand control productivity. Flow Control Devices (FCDs) have demonstrated significant potential for improving recovery in SAGD production wells. FCD experience in SAGD has been primarily positive and most producers performed better with FCDs. While for some operators the results are mixed or negative. The first generation of FCDs deployed in SAGD projects were not designed for such process and in some cases ill-suited. Furthermore, current modeling methods deployed have significant limitations that prevent appropriate FCD design or understanding, and although they can be history matched, typically do not provide useful insight into the pressure drops encountered or held understand the benefit of different devices. In order to design and optimize FCDs for SAGD, it is necessary to characterize different FCDs under the steam-breakthrough condition, and accurately model the flashing in the near wellbore area associated with low-subcool operation. The extensive chocking in FCDs, far greater than initial design, in many cases is due to near-wellbore flashing. This work is a continuation of three previous parts discussing the liquid pool modeling for SAGD producers ( Irani, 2018 , 2019 and Irani and Gates, 2018 ). The purpose of this work is to create a PI that fit for purpose of SAGD liquid pool pre- and post-flashing that mainly can be used for analysis and optimization of FCDs. With FCDs, the draw-down pressure is typically higher, resulting in flashing near the well bore. If there is flashing in the near wellbore area, the temperature gradient within liquid pool yields the saturation curve. The flashing causes the reduction in the relative permeability of the liquid phase, that creates new equilibrium that stabilizes at lower rates. Such new equilibrium analysis is conducted by forcing a new temperature gradient to the model. The main output of such analysis is the produced steam quality at the producer sand-face. The steam quality is an important input for the flow control devices (FCDs) especially at subcool close to the zero, as it controls its behavior. This type of analysis can help the operators evaluate the effectiveness of different type of FCDs, whether they are primarily momentum- or friction-style devices.

Chemical additives for co-injection with steam at low dosages were evaluated for enhancement of SAGD operations. The additives were selected based on their commercial scalability, water solubility and transportability inside the SAGD chamber (volatility). A custom-designed oil-recovery experiment, mimicking actual mechanisms of a SAGD chamber edge, was utilized for additive evaluation in the laboratory. The apparatus is a steam-soaking system composed of a synthetically prepared oil sand core suspended above an additive solution. The core was internally cooled to drive steam condensation and create co-current oil/water drainage inside the core. The additive effect on the recovery rate was quantified by comparison with a steam-only baseline. Amongst 20 additives tested, Additives A and B showed the most promising performance. Additive A improved the rate of oil recovery by 18% at a dosage of 5000 ppm. When the dosage was doubled (1 wt %), improvement in the rate of recovery increased favorably by 5 times (93% over baseline). Additive B showed significantly higher oil recovery of 240% over the baseline at 5000 ppm. The transport behaviors of Additives A and B inside the steam chamber were also evaluated with reservoir modeling for potential field implementation. A homogeneous reservoir model was used to predict the transport behavior of the additives in a growing chamber and evaluate multiple injection scenarios. Compared with Additive A, Additive B was less volatile and required longer travel time from the injector to the chamber edge. Hence, Additive B can be injected in a younger (smaller) steam chamber, where the chamber edge is more accessible by the additive. Additive A, on the other hand, can target regions away from the injector in a more developed steam chamber. Additives A and B could also be co-injected or alternatively injected regardless of chamber size to simultaneously enhance oil production from different regions of the chamber.

Multiphase relative permeability is a key parameter in reservoir simulation. Typically, end-point based correlations are employed in order to obtain such curves for reservoir simulation purposes. However, those correlations are not capable of capturing micro-scale physical phenomenon which can significantly affect flow pattern at larger scales. Consequently, it is necessary to obtain a scale-up methodology in order to transfer the micro-scale physics to reservoir-scale. The objective of this research is developing a scale up procedure which can be applied to multi-phase flow properties obtained by micro-scale flow simulation to compute the equivalent macro-scale and core-scale flow properties having the micro-scale flow properties. Two different sets of media are employed: media representing unconsolidated oil sands and media based on experimental data obtained from the Mesaverde formation located in the Poweder River Basin. The former is used for validating the scale up methodology since not all the required information is available in the experimental data set. Pore scale network modelling is used for calculating micro-scale multi-phase flow properties such as porosity, and absolute and relative permeability. Then the generated subsegments are populated in space to reconstruct the macro scale medium. Flow properties of such medium are then obtained by network modelling and the proposed scale-up methodology and the results are compared. Furthermore, macro-scale media are distributed in space in layers and stacks of increasing and decreasing permeability to form a core-level medium. Single and multi-phase flow properties are then calculated by applying a pressure drop across the core. Permeability and relative permeability curves are calculated using the combination of mass balance, equation of state, and Darcy equation assuming steady-state flow while capillary pressure curve is obtained using the modified Leverett-J function procedure used in micro-to-macro scale up section. Results show good agreement between the expected and calculated properties for both unconsolidated and consolidated media. Finally, physical behavior observed at micro and macro scale is transferred to the core scale.

Steam-assisted gravity drainage (SAGD) recovery process is strongly impacted by the spatial distributions of heterogeneous shale barriers. Though detailed compositional flow simulators are available for SAGD recovery performance evaluation, the simulation process is usually quite computationally demanding, rendering their use over a large number of reservoir models for assessing the impacts of heterogeneity (uncertainties) to be impractical. In recent years, data-driven proxy has been widely proposed to reduce the computational effort; nevertheless, the proxy must be trained using a large data set consisting of many flow simulation cases that is, ideally, spanning the model parameter spaces. The question remains: is there a more efficient way to quickly screen a large number of heterogeneous SAGD models? Such techniques could help to construct a training data set with less redundancy; they can also be used to quickly identify a subset of heterogeneous models for detailed flow simulation. In this work, two particular distance measures, flow-based and static-based, are formulated to quantify the similarity among a set of 3D heterogeneous SAGD models. First, to formulate the flow-based distance measure, a physics-based particle tracking model is employed: Darcy's law and energy balance are integrated to mimic the steam chamber expansion process; steam particles that are located at the edge of the chamber would release their energy to the surrounding cold bitumen, while detailed fluid displacements are not explicitly simulated. The steam chamber evolution is modeled, and a flow-based distance between two given reservoir models is defined as the difference in their chamber sizes over time. Second, to formulate the static-based distance, the Hausdorff distance is employed: it is often used in image processing to compare two images according to their corresponding spatial arrangement and shapes of various objects. A suite of 3D models is constructed using representative petrophysical properties and operating constraints extracted from several pads in Suncor's Firebag project. The computed distance measures are used to partition the models into different groups. To establish a baseline for comparison, flow simulations are performed on these models to predict the actual chamber evolution and production profiles. The grouping results according to the proposed flow- and static-based distance measures match reasonably well to those obtained from detailed flow simulations. Significant improvement in computational efficiency is achieved with the proposed techniques. They can be used to efficiently screen a large number of reservoir models and facilitate the classification of these models into groups with distinct shale heterogeneity characteristics. It presents a significant potential to be integrated with other data-driven approaches for reducing the computational load typically associated with detailed flow simulations involving multiple heterogeneous reservoir realizations.

Since decades, steam-assisted oil recovery processes have been successfully deployed in heavy oil reservoirs to extract bitumen/heavy oil. Current resource allocation practices mostly involve reservoir model-based open loop optimization at the planning stage and its periodic recurrence. However, such decades-old strategies need a complete overhaul as they ignore dynamic changes in reservoir conditions and surface facilities, ultimately rendering heavy oil production economically unsustainable in the low-oil-price environment. Since steam supply costs account for more than 50% of total operating costs, a data-driven strategy that transforms the data available from various sensors into meaningful steam allocation decisions requires further attention. In this research, we propose a purely data-driven algorithm that maximizes the economic objective function by allocating an optimal amount of steam to different well pads. The method primarily constitutes two components: forecasting and nonlinear optimization. A dynamic model is used to relate different variables in historical field data that were measured at regular time intervals and can be used to compute economic performance indicators (EPI). The variables in the model are cumulative in nature since they can represent the temporal changes in reservoir conditions. Accurate prediction of EPI is ensured by retraining the regression model using the latest available data. Then, predicted EPI is optimized using a nonlinear optimization algorithm subject to amplitude and rate saturation constraints on decision variables i.e., the amount of steam allocated to each well pad. The proposed steam allocation strategy is tested on 2 well pads (each containing 10 wells) of an oil sands reservoir located near Fort McMurray in Alberta, Canada. After an exploratory analysis of production history, an output error (OE) model is built between logarithmically transformed cumulative steam injection and cumulative oil production for each well pad. Commonly used net-present-value (NPV) is considered as EPI to be maximized. Optimization of the objective function is subject to distinct operating conditions and realistic constraints. By comparing results with field production history, it can be observed that optimum steam injection profiles for both well pads are significantly different than that of a field. In fact, the proposed algorithm provides smooth and consistent steam injection rates, unlike field injection history. Also, the lower steam-oil ratio is achieved for both well pads, ultimately translating into ∼19 % higher NPV when compared with field data. Inspired from state-of-the-art control techniques, the proposed steam allocation algorithm provides a generic data-driven framework that can consider any number of well pads, EPIs, and amount of past data. It is computationally inexpensive as no numerical simulations are required. Overall, it can potentially reduce the energy required to extract heavy oil and increase the revenue while inflicting no additional capital cost and reducing greenhouse gas emissions.

The simulation of thermal recovery processes and their variations like non-condensable gases (NCG) co-injection with steam in SAGD require reliable gas-liquid relative permeability curves. In this work, three experiments were conducted to extract equivalent thermal relative permeability curves, and the experimental curves were used as an input in a numerical simulation model to evaluate the NCG co-injection with steam in SAGD. Heavy oil was diluted with toluene to mimic the viscosity of the oil at 165, 195, and 225°C. It was assumed that the viscosity variation was the main factor affecting the temperature-dependent gas-liquid relative permeability curves. Athabasca sand was packed, and after water and oil were flooded, gas was injected. Three sets of gas-liquid relative permeability curves were extracted through history match. The experimental curves were used as an input in a simulation model to evaluate the effect of the amount of NCG and time of NCG co-injection in SAGD using CMG STARS. In contrast to previous simulation studies, the current simulation results showed that most of the injected gas leaked from the steam chamber and accumulated at the top of the reservoir. Part of the gas stayed at the top of the steam chamber insulating the steam chamber from the overburden rock. The gas that leaked from the steam chamber aided in additional recovery of oil. According to the simulation results, for a homogeneous reservoir, small amounts of NCG (about 1 mole %) can be co-injected with steam at any time of a SAGD project to reduce its energy requirements without a negative impact in the oil recovery. This amount can be progressively increased to further reduce the energy requirements and increase the ultimate oil recovery. With the introduction of the new experimental gas-liquid relative permeability curves, it would be possible to optimize NCG co-injection processes in SAGD with the use of numerical simulation, taking into account the flow behaviour of gases in the reservoir that usually is not considered.

Steam Assisted Gravity Drainage (SAGD) has been widely applied to unlock hydrocarbon resources in oil sands reservoirs. This method uses steam, which is generated at the surface, to heat a formation and create a steam chamber around an injector. Past studies have indicated that reservoir heterogeneity is one of the crucial factors that directly affectthe performance of the SAGD process. This paper presents an innovative integrated modeling approach for evaluating, assisted history matching, and production forecasting of the SAGD process with the presence of a complex shale barrier system in oil sands reservoirs. As SAGD is a strongly geological dependent recovery process and, unfortunately, there are many uncertainties associated with reservoir geology in reality. Therefore, it requires generating a large number of geological realizations to capture the critical effects of geology, especially with the presence of shale barriers, in history matching and field development planning of the SAGD process. To quantify the impact and improve the quality of history matching compared with the traditional method, an efficient integrated workflow has been developed in which geological information generated from a geological modeling package is automatically updated for a reservoir simulator and controlled by an intelligent optimizer in a big-loop modeling approach. A detailed workflow on the integrated modeling approach that includes shale barriers for a typical oil sands reservoir is described in the first section of this paper. Shale bodies are geostatistically distributed in the geological models. A comprehensive parametric study was conducted with numerous geological realizations to identify the critical role of shale barriers in SAGD performance including shale geometry, shale length and thickness, shale distribution and proportions. Then the Bayesian algorithm with a Proxy-based Acceptance-Rejection sampling method is employed for assisted history matching of SAGD production profiles. With the presence of complex shale barriers, it requires simultaneous updating of both geological and reservoir engineering parameters. Using the proposed approach, the global history matching errors were drastically reduced in all production wells. Validation results indicate that the integrated modeling approach effectively helps to update the properties and distribution of shale barriers to find the closest geological distribution compared to the true solution. Finally, an ensemble of the best-matched simulation models is used to perform a probabilistic forecasting to capture the uncertainties in future production profiles. Not limited to history matching ofthe SAGD process, the proposed approach can be also applied to different complex problems such as robust optimization for various recovery methods from conventional to unconventional reservoirs.

Recently, coupled electromagnetic-reservoir simulators based on full numerical schemes have been developed using commercial software packages. However, simplified analytical and semi-analytical models can provide useful and quick insight of the complicated coupling between radio wave propagation and transport phenomena in heavy oil reservoirs. In this study, we developed simplified coupled EM-thermal models to emphasize on two important aspects of radio frequency heating process, near-field analysis and insulated antenna setups. It will be shown how much thermal error can be introduced when a proper full-wave analysis is not taken into account. We will also demonstrate the thermal advantage of applying an insulated antenna over a bare antenna when radio waves are used for heating purposes. Another key aspect of radio-frequency heating process is to consider dependency of electrical properties to thermal and water saturation conditions. This one is more difficult to verify analytically and it will be done in the second part of this study.

Impacts of reservoir heterogeneities in the form of shale barriers on SAGD production can be analyzed by generating a large number of realizations of shale barrier configurations and subjecting them to flow simulation. However, visualizing and quantifying the (dis)similarities among these realizations is often challenging. A workflow that applies multidimensional scaling (MDS) and cluster analysis techniques is developed to represent the uncertain influences of different shale barrier configurations on SAGD production and to quantify the dissimilarities between realizations. A two-dimensional homogeneous simulation model is employed, and reservoir heterogeneities are simulated by superimposing sets of idealized shale barriers on the homogeneous model. The petrophysical properties, such as the porosity, permeability, initial oil saturation and net pay thickness, have been taken from average values for several pads in Suncor's Firebag project. One thousand models with various shale barrier configurations are then subjected to flow simulation to estimate SAGD production in each case. First, a distance function, which measures the dissimilarity in production responses between any two given shale barrier configurations, is formulated. Next, MDS maps the resultant distance matrix into an n-dimensional Euclidean space, where k-means clustering technique is applied to group the models into multiple clusters. Although the precise distribution of shale barriers would vary among models within the same cluster, it is expected that their impacts on SAGD production are similar. Specific features corresponding to the shale barriers in each cluster are analyzed, and they are studied to infer any potential correlation between SAGD production and the particular shale distribution characteristics. The results are employed to revise the original set of realizations by adding new models to clusters with fewer members and removing models from clusters with redundant members. The new models are subjected to flow simulation to verify their membership to the assigned clusters, and good agreement in the results has been observed. Data-driven or AI-based modeling approaches for production analysis have gained much attention over recent years. In most cases, a training data set consisting of many different realizations of reservoir heterogeneity is needed. A key question remains: "how many realizations are needed to span the model parameter space?" The proposed workflow offers an efficient and systematic method for constructing data sets that maximize the spanning of the model parameter space, without exhaustively sampling similar realizations and subjecting them to flow simulation. This is a particularly important consideration when 3D models are utilized. Furthermore, the ability to visualize and select representative models or scenarios from individual clusters has important potential for facilitating improvements in operations design in the presence of reservoir heterogeneities.