Cyclic solvent injection (CSI) is a promising technology for enhanced production of heavy oil in post-CHOPS (Cold Heavy Oil Production with Sand) fields. The first stage of a CSI cycle involves the injection of solvent vapour, which re-energizes the reservoir and dissolves into oil, reducing its viscosity. The second stage of a CSI cycle involves dropping the pressure to flow the solvent-diluted oil back to the well. This is the solvent analog to cyclic steam stimulation (CSS). A key difference between solvent and steam is that the heat from steam stays in the oil even during production, while in solvent-diluted oil, if solvent comes out of solution as pressure drops, oil viscosity will increase again and one of the main benefits of the solvent will be lost. Solvent selection in CSI needs to consider which solvents have non-equilibrium properties, specifically delayed release during pressure depletion.
This study presents a set of low field nuclear magnetic resonance (NMR) tests that were run on a 3,370 mPa·s viscosity heavy oil and several solvents: methane-propane and methane-CO2. NMR measures the relaxation rate of protons in the presence of magnetic fields. When fluid viscosity decreases, the fluid signal relaxes more slowly. A properly calibrated NMR model can be used to measure the viscosity of the liquid phase in a solvent-oil mixture as pressure drops and gas leaves solution. This approach does not require flow, and can also be used to measure the
Tests were run using methane-CO2 solvents with varying CO2 concentration to study the impact of live oil viscosity on the rate of solvent release. For lower initial live oil viscosity, the solvent is able to more quickly leave solution from the oil. Additional tests run using a methane-propane solvent demonstrate the impact of varying solvent type for the same initial solution viscosity. Both CO2 and propane show significant non-equilibrium solvent release during depletion.
Measurements of non-equilibrium solvent release from oil are an important piece in the understanding and modeling of heavy oil CSI. Equilibrium PVT data shows how different solvents can affect heavy oil systems, but the non-equilibrium solvent release is crucial for these solvents to work as recovery agents in the field. The results presented in this study provide useful data for CO2 and propane as potential CSI solvents, and help in the understanding of the role of viscosity vs. solvent type in the diluted oil response.
Liu, Hao (China University of Petroleum) | Cheng, Linsong (China University of Petroleum) | Huang, Shijun (China University of Petroleum) | Chen, Xiao (China University of Petroleum) | Huang, Qian (China University of Petroleum)
Co-injection of solvent and steam in Expanding Solvent-Steam Assisted Gravity Drainage (ES-SAGD) increases the mobility of highly viscous oil relative to conventional SAGD wherein only steam is injected. This consequently results in higher oil production rates and lower energy consumption relative to SAGD due to the combined benefits of heating and dilution. Current efforts have been made to study mass transfer from a single-component-solvent/steam mixture into the heavy oil at the steam front, while the low-cost multicomponent solvent is actually co-injected with steam on the field, which presents a much more complicated interfacial condition.
In this paper, a generalized methodology has been developed to couple heat and mass transfer of this solvents–steam–heavy oil system at the interface. In the model, the equation of state and fugacity equation are integrated to calculate and analysis the equilibrium state of transition interface. Then, the transition interface model is established based on heat transfer formulations and gaseous phase convective equations. Finally, under the temperature and velocity boundaries, the model is solved by applying implicit Runge-Kutta method and the results are used to analysis the heat and mass transfer characteristics.
The predicted results show that increasing the fraction of lighter solvent can enhance mass transfer, while the heat transferred is seriously deteriorated. Besides, the closer the bubble point temperature and dew point temperature of the stream is, the smaller the elevation of mass transfer will be, but the heat transfer is almost unaffected. Therefore, to realize the best effects of heat and mass transfer, solvents with larger saturation temperature difference should be applied, and the composition of the solvents should be carefully determined according to the formation properties and solvent components.
Conducting an ES-SAGD project on a field has many challenges, especially considering its main economic factors: production increase and solvent cost. With the new methodology, an optimum multicomponent solvent can be screened from the ocean of solvent candidates for a best heat and mass transfer performance in an ES-SAGD process and hence the oil flux at the chamber edge can be maximized with a minimized solvent cost. Therefore, it is of fundamental and practical importance to study the coupled heat and mass transfer of multicomponent solvents–steam–heavy oil system at the interface.
Khaledi, Rahman (Imperial Oil Resources) | Motahhari, Hamed Reza (Imperial Oil Resources) | Boone, Thomas J. (ExxonMobil Retiree) | Fang, Chen (ExxonMobil Upstream Research Company) | Coutee, Adam S. (ExxonMobil Upstream Research Company)
Thermal-Solvent Assisted Gravity Drainage processes are heavy oil recovery processes in which the stimulation mechanism for bitumen viscosity reduction is by heating and dilution. The range of the injected solvent concentration with steam may be low such as in Solvent Assisted-SAGD or very high such as in Heated Vapour Extraction. The performance behaviour of these processes is significantly driven by the complex thermodynamic interaction of steam and solvent, heat transfer, multiphase fluid equilibrium and flow in the porous medium.
ExxonMobil and its affiliate Imperial have been optimizing the existing recovery processes and developing new technologies to improve the efficiencies and environmental performance of the heavy oil production operations. Recent focus has been on developing thermal-solvent based recovery processes through an integrated research program that includes fundamental laboratory work, advanced numerical simulation studies, laboratory scaled physical modeling, and field piloting. The research program aims at in-depth investigation and understanding of process physics and mechanisms, and evaluating process performance and behaviour to enable development of new recovery methods and to enhance the performance.
This paper focuses on the fundamental concepts of Azeotropic Heated Vapour Extraction, a new thermal solvent recovery technology developed by ExxonMobil-Imperial. This technology takes the combined advantages of the solvent dilution mechanism for enhanced oil production rate with the minimum required energy (GHG emission), as well as the effectiveness of energy transport of steam to minimize the required solvent in circulation for the extraction process. In this paper, the complex solvent-steam phase behaviour and reservoir fluid flow and their interaction under operating conditions are investigated. The analysis of experimental and modeling data shows that the injection of solvent-steam mixture at its azeotropic condition results in significant improvement in process key performance indicators (KPI's). At these conditions, the reservoir is heated to the minimum boiling temperature of the solvent-steam mixture compared to a Heated VAPEX or Solvent Assisted-SAGD process resulting in the reduction of the required energy thereby minimizing the solvent-to-oil ratio. Also, due to phase equilibrium, the vaporization of in-situ water is prevented resulting in the reduction of retained solvent in the depleted zone of the reservoir. It is found that an improvement in the process KPI's is dependent on the volatility of the selected solvent. The process KPI's also vary with operational conditions. The recovery process is optimized for certain reservoir constraints through the selection of the solvent boiling range and consequently the azeotropic steam content in the injected mixture.
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.
The CHOPS process involves the growth of high permeability channels (wormholes) into the reservoir. The wormholes provide improved access to the reservoir, thereby substantially increasing oil production rates. Field evidence and laboratory experiments indicate that wormholes tend to persist as stable high-permeability flow channels throughout the duration of the CHOPS process.
The existence of wormholes in the reservoir needs to be taken into consideration in the design of post CHOPS recovery processes. One of the key issues is the stability of wormholes during the re-pressurization phase of any follow up process to CHOPS. For example, wormholes could remain as stable high permeability flow channels to deliver injected fluids into a reservoir or they could collapse to improve conformance for fluid injection.
This paper presents the results of experiments testing the stability of wormholes under conditions that could be destabilizing such as cold and hot water circulation, heating, and pressurization with hydrocarbon gases (e.g., methane, ethane). The experiments were performed in a triaxial cell, approximately 19 cm in diameter and 18 cm long, packed with water-wet sand and saturated with heavy oil. Wormholes were generated in the pack by injecting heavy (dead) oil into the cell at a sufficiently high rate to produce sand from the pack. Then, conditions that could be destabilizing to wormholes were applied to test their outcome.
Cold water flowing through the wormhole eroded its walls, eventually causing the pack to collapse after more than a week. The injection of hot water accelerated the process of erosion considerably. Heating the pack did not cause it to collapse, but the wall of the wormhole was destabilized to a limited extent in that the diameter of the wormhole increased by more than a factor of two. The heating reduced the viscosity of the heavy oil in the pack, causing the apparent strength of the sand to decrease. Pressurizing the sand pack with methane and then slowly depressurizing it did not destabilize the wormhole. However, pressurizing the pack with ethane caused the sand matrix surrounding the wormhole to collapse over a period of two weeks. As in the case of heating, dissolution of ethane into the heavy oil caused the apparent strength of the sand to decrease by reducing the viscosity of the heavy oil in the pack.
Although the confining pressure in these experiments was relatively low, the experimental cell did not allow for stress redistribution due to the deformation of the sand matrix, as would occur in the field. For this reason, the wormholes generated in the experiments were less stable than ones in the field. Consequently, the laboratory conditions that resulted in wormhole collapse should be regarded as a guide to the behaviour that would be expected in the field.
In in-situ oil sands operations using Steam Assisted Gravity Drainage (SAGD), understanding the temperature and pressure responses during periods of shut-in and slow-down is crucial for effective steam chamber management. The corresponding operating strategy should be based on the temperature and pressure performance.
At Nexen’s Long Lake in-situ project, there are a number of pads with extensive top water in contact with the target bitumen interval. Along with the rest of the wells in operation at Long Lake, these SAGD wells have experienced several periods of shut-in and slow-down due to maintenance operations and other surface constraints. Surface and downhole sensors and other real-time monitoring technologies provide opportunities to monitor temperature and pressure during these time periods, which help to better understand the impact of surface activities on steam chamber development, particularly in areas with extensive top water present.
This paper presents the temperature and pressure responses over time at different elevations within the reservoir. It compares temperature and pressure responses for the wells with top water and without top water and 4D seismic is also integrated to analyze the causes for the different temperature and pressure responses. Review of the data shows that shut-in has a more significant impact on temperature and pressure in the steam chamber than slow-down. Wells with top water experience severe temperature drops while slow-down wells do not. After long periods of shut-in, areas with thicker top water may take additional time to heat up and reach steam conditions again, resulting in higher steam injection requirements. The collected data suggests that continuous steam injection is crucial to maintain steam chamber with top water.
Thermal-Solvent Assisted Gravity Drainage are recovery processes in which the stimulation mechanism for bitumen viscosity reduction is by heating and/or dilution. Different gravity drainage recovery processes can be included in this category depending on the range of the injection temperature and the solvent concentration in the injection stream. Examples are SAGD, SA-SAGD, VAPEX and heated VAPEX. The performance behavior of these processes is significantly driven by the complex thermodynamic interaction of steam and solvent, heat transfer, multiphase fluid equilibrium and flow in the porous medium.
In this study, we develop a general analytical model for gravity drainage processes by incorporating mass transfer mechanisms (including diffusion and dispersion) and heat transfer mechanism by conduction. In particular, we incorporate the dependency of diffusion and dispersion coefficients on concentration, temperature and drainage velocities, respectively. We utilize a novel approach to analytically solve the second order non-linear partial differential equation which governs mass transfer within the mass boundary layer. The resulted closed-form analytical model provides oil drainage rate due to gravity and heat and dilution effects as a function of reservoir and fluid properties. The developed model in this work provides a new perspective into the mass transfer mechanisms and their relative importance within the mass transfer boundary layer at different operating conditions.
The consistent application of the new model to the gravity drainage processes ranging from SAGD to VAPEX demonstrated using laboratory data from literature. It is shown that the predicted concentration distribution profile by the model with the concentration-dependent diffusion coefficient is profoundly different than the predicted profiles with a constant diffusion coefficient. The modeling results demonstrate that the dispersion can be several orders of magnitude greater than diffusion for solvent assisted gravity drainage process at the elevated temperatures. In addition, the contribution of the mass transfer boundary layer to oil production rate can be significantly greater than the heat transfer boundary layer despite being considerably narrower than the heat transfer boundary layer at the elevated temperatures. These findings confirms that the performance of solvent assisted gravity drainage process can be more favorable at the elevated temperatures.
The aim of this work is to evaluate the micro-displacement mechanisms associated with heavy oil recovery by water and chemical flooding in a non-linear system (2D Model). To evaluate the effect of sweep efficiency improvements with no local pore level trapping of fluids, a 2D neutral water wet Hele-Shaw cell is used to visualize the dominant mechanisms of surfactant, polymer and their combinations to enhance heavy oil recovery. Waterfloods were conducted as a base line, and compared to surfactant, polymer and SP floods at different injection stages (secondary and tertiary) and under variable injection rates. Post-breakthrough oil recovery and pressure drop during the water flood and chemical flood were evaluated with the analysis of images which show the distribution of the fluids in the cell.
The results demonstrated that in waterflooding, oil pinches off the pre-formed water fingers and allows the formation of snap-off which divert further water to un-swept zones. Oil is produced discontinuously because of snap-off effects, with production of oil related to small pressure spikes in the system. This mechanism was also observed in polymer displacements. This observation can explain viscous flow of oil and discontinuous oil and water production after breakthrough in a heavy oil waterflood. Oil stripping and zones of low IFT, in the form of O/W emulsions were mechanisms observed in surfactant and SP flooding. Such emulsions are present in secondary fingers that then propagate and sweep new areas of the cell. The synergy between polymer and surfactant leads to a slightly better sweep efficiency in the cell by stripping oil and stabilizing the flood front compared to surfactant or polymer alone. Injection rate has also an impact in the flow stability and formation and flow of emulsions. At higher injection rates, flow is more unstable and leads to a less oil recovery. Multiple tests under the same conditions revealed the reproducibility levels of each displacement process.
The novelty of this study is that it provides insights into fluid flow behavior in diverging flow paths as observed in 2D systems, as opposed to linear core floods that have limited flow pathways. In particular, these visual observations help to illustrate the role that surfactants can play in heavy oil systems, with and without the addition of polymer.
Going forward, limiting greenhouse gas emissions and reducing water usage in oil sands upstream operations is essential to meeting regulatory requirements and a pre-condition for the social license to operate. As Oil Sands Industry exhausts its best quality rich pay and gradually move into producing bitumen from marginal quality assets with thin pay zones, SAGD performance will deteriorate with higher steam to oil ratios. In this paper, Enhanced Convective Solvent Extraction (ECSOLVEX) process is described to apply to such reservoirs.
ECSOLVEX predominantly uses a pure solvent which is injected into the reservoir at or above its dew point temperature. Injection and production wells are placed some distance apart in the same lateral plane. Sidetracks from injector well to close proximity of the producer well allow a solvent chamber to extend to the inter-well region. The process is primarily gravity driven with a mild convective gradient from injector to the producer well. A solvent chamber is established along the side track lateral, which grows parallel to the injector and producer wells. Water production is limited since it is a solvent driven processthat would result in a reduced footprint for the surface oil water separation facility.
Simulation results show that each side track creates its own drainage chamber which performs independently of other drainage chambers till they merge, thus the production performance of this process is a function of the number of sidetracks drilled. As an example, with propane as the solvent, production rates can be doubled that of SAGD if sidetracks are drilled every 100m. The energy consumption and associated emissions of such a process would be 70% less than SAGD. The performance can be further improved by localized heating in the proximity of the producer well.
ECSOLVEX process concept provides the opportunity for economic recovery from pay zones which are not thick enough for traditional SAGD process to operate. It demonstrates significantly lower energy consumption as compared to SAGD, leading to a lower carbon footprint and has the potential to reduce the size of the oil water separation facility.
A generalized model is presented to calculate the k-values of methane/bitumen, ethane/bitumen, propane/bitumen, and butane/bitumen systems. These data are required for phase behaviour modeling and simulation of solvent-aided bitumen recovery processes. The proposed model is evaluated by comparing the calculated results with the measured experimental k-values.
The proposed model provides generalized binary interaction parameters between each hydrocarbon solvent (methane, ethane, propane, and butane) and the defined components in bitumen and calculates the k-values of solvent/bitumen systems. Unlike the existing common approaches, experimental solubility data are not required to tune the model. The boiling point or carbon number distribution of bitumen or heavy oil obtained by simulated distillation (SimDist) test is the only required data to characterize and define the components of heavy oil or bitumen. The SimDist test is a very fast test and much less expensive than the common solubility measurements.
This model has been developed based on the experimental fractionation of bitumen. The Athabasca bitumen was experimentally fractioned to four bitumen cuts applying vacuum distillation method and the solubility of solvent in each bitumen cut were measured at wide ranges of temperature and pressure. The measured solubility data of methane, ethane, propane, and butane in each bitumen cut have been used to tune the PR-EoS and the generalized binary interaction parameter coefficients for each solvent and bitumen components have been found. To calculate the k-values of solvent/bitumen mixtures, the bitumen is defined as a mixture of n-alkanes based on simulated distillation results. The properties of n-alkanes have been assumed for each component. Employing the obtained binary interaction parameters in PR-EoS using experimental data of solvent/bitumen cut systems and considering the defined bitumen components as input to the proposed model, the k-values of solvent and any bitumen or heavy oil mixtures are calculated. The validity of the proposed model has been confirmed by calculating the k-values of methane, ethane, propane, and butane with two different bitumen samples with an average deviation of less than 3.0 %.
The proposed model predicts the k-values of methane/bitumen, ethane/bitumen, propane/bitumen, and butane/bitumen mixtures without requiring the time and cost intensive experimental PVT data for tuning. Providing the simulated distillation results of any bitumen or heavy oil samples, the k-values of hydrocarbon solvent/bitumen or heavy oil can be calculated in wide ranges of temperature and pressure. The outputs of this model can be directly used as k-values in simulation of solvent-aided thermal recovery processes.