At the Nexen Long Lake in situ steam-assisted gravity drainage (SAGD) oil sands recovery project, the bitumen-saturated reservoir is in the Lower Cretaceous McMurray Formation. The main depositional environment in the reservoir unit is fluvial-estuarine meandering channels. Stacked channel deposition exhibits a high degree of variability both vertically and laterally over short distances and depositional complexity occurs at many scales. Many papers have been written on characterizing oil sands deposition geologically or geostatistically. However, complete characterization cannot be achieved at all scales due to the degree of complexity.
Building a history-matched geomodel can be very time consuming and very challenging in complex reservoirs such as in Long Lake, where the Quaternary (Gregoire) Channel, collapse features, top gas and top water, lean zones, as well as shale barriers and baffles, contribute to the complexity. This paper presents a practical geological modeling approach used at Nexen to quantify uncertainties of reservoir properties. This approach has been validated by history matching and prediction. The solution is based on the integration of all available geology, geophysics, petrophysics, reservoir engineering, and production information. Using the proposed solution, the number of modeling iterations and the time required to achieve the desired objectives of history matching and prediction have been significantly reduced.
Crude bitumen extracted from the Canadian oil sands has a high viscosity, so typically it does not flow at normal pipeline temperatures. After extraction the bitumen can be mixed with a diluent (refined naphtha, condensate or SCO) before being pumped by pipeline to a refinery for processing. Alternatively the bitumen can be processed to either improve the viscosity, produce a higher value synthetic crude oil (SCO), or fully finished products before export. These additional costs can significantly impact the overall economics of bitumen extraction.
The Veba Combi-CrackingTM (VCCTM) process is a proven slurry based hydrogen addition technology able to process refinery residues, heavy crude oil and even coal. Commonly found in a refinery environment it is also ideally suited to field upgrading, since it can be tailored to produce directly marketable products, a high quality synthetic crude oil, or simply provide viscosity reduction for easy transportation.
This paper shows that the VCCTM technology is ideally suited to field upgrading of bitumen. With an Athabasca Bitumen feedstock the VCCTM process can achieve greater than 90% conversion of asphaltenes, and a 95% overall conversion of the 524oC+ material to desirable products, in a single pass. This paper also shows an improvement in overall efficiency and minimisation of undesirable products when compared to other common field upgrading processes. With simple changes to the flowsheet it is capable of producing high quality finished products (Naphtha: 1 ppm S, Diesel: <10 ppm S, Cetane >45, VGO: <100 ppm S, Metals <1 ppm), a fully upgraded syncrude, or alternatively provide significant viscosity reduction (example heavy crude VR reduced from 145,800 cSt to 3.0 cSt) for export by pipeline.
In thermal compositional reservoir simulators that use energy as a primary variable, thousands to millions of isenthalpic multiphase flash calculations must be performed to calculate temperature, phase splits and compositions for different grid blocks during the simulation. Development of a robust and fast isenthalpic multiphase flash calculation method is necessary to improve the efficiency of such simulations.
A new isenthalpic multiphase flash calculation is described in this paper. The flash calculation method uses a modified Rachford-Rice monotonic objective function and the negative flash concept for phase distribution and phase identification. Therefore phase stability analysis is not necessary. The formulation and algorithm of the new method are presented in detail. This method is able to handle difficult situations such as narrow boiling point regions and phase appearance and disappearance, which are dominant in thermal processes. The current method encounters no difficulty in the latter situations unlike stage-wise isenthalpic flash calculation methods.
After the accuracy of new method was compared and verified against current algorithms used by the industry, it was also tested for robustness and speed. The results show promising performance compared to the current methods. This proposed method is not sensitive to the initial guess for temperature. As a matter for fact, in all of the test cases in this study, the same temperature was used as the initial guess. A poor initial guess for temperature only requires more iterations to reach the solution.
Inflow control devices (ICDs) have been extensively used in horizontal wells for conventional oil and gas production in order to prevent early water break through or gas coning. The benefits associated with this technology have been studied with reservoir simulation and validated with field experience. Some of the benefits associated with ICDs that have been described in the literature are: easier well clean out during start up because the ICDs allow application of higher drawdown to poorly performing sections of the reservoir, higher recovery factor caused by delayed water break through or gas coning, uniform production contribution of the horizontal section, and better sand control by limiting the fluid rate per joint.
In theory, similar benefits can be achieved using ICDs in SAGD applications. In this case steam break through instead of gas coning could be prevented. If live steam production is mechanically controlled by the ICDs, bitumen from the colder sections of the production well can be produced by applying a higher drawdown. In this way the total bitumen production should be increased while steam oil ratio (SOR) is reduced. Nevertheless, the available drawdown between SAGD well pairs is very limited and the conventional design approach consisting of uniform flow along the horizontal section is not applicable.
To design an ICD completion trial at the MacKay River SAGD Project, Brion Energy has conducted studies involving theoretical analysis, segmented well simulation, technology evaluation, and equipment selection. The results of these studies indicate that higher cumulative production with lower SOR can be achieved when the ICDs are installed in bitumen production wells where the reservoir exhibits significant heterogeneities. However, the hot spots in the well have to be produced at zero wellbore sub-cool in order to generate a differential pressure capable of bringing colder sections of the well into production. The study also suggests that a further reduction in SOR can be achieved when the ICDs are also installed in the steam injection wells.
Desheng, M. (Research Institute Of Petroleum Exploration & Development) | Lanxiang, S. (Research Institute of Petroleum Exploration & Development) | Changfeng, X. (Research Institute Of Petroleum Exploration & Development) | Xiuluan, Li (Research Institute Of Petroleum Exploration & Development) | Erpeng, Guo (Research Institute Of Petroleum Exploration & Development) | Fengxiang, Yang (Research Institute of Petroleum Exploration & Development, PetroChina Xinjiang Oilfield Company) | Xiaorong, Shi (Research Institute of Petroleum Exploration & Development, PetroChina Xinjiang Oilfield Company) | Yaoli, Shi (Research Institute of Petroleum Exploration & Development, PetroChina Xinjiang Oilfield Company) | Changjun, Diao (Research Institute of Petroleum Exploration & Development, PetroChina Xinjiang Oilfield Company)
Steam assisted gravity drainage (SAGD) enjoys great advantages in the development of extra heavy oil reservoir such as high oil rate and favarable oil steam ratio. However, there are also disadvantages, such as intensive energy consumption, produced water recycle and disposal, that have impacts on the economics of SAGD projects. Furthermore, oil steam ratio declines and water cut rises when the SAGD comes into its later stage while great residual oil existing in the wedge zone. This paper proposes a new method, turning to in situ combustion (ISC) in the later SAGD, to improve the performance of the later SAGD. The feasibility and performance are both studied systematically. Firstly, a numerical model is established on the basis of reservoir and fluid parameters from a block in Xinjiang oil field, China, and then the performance characteristics in different stages of SAGD in extra heavy oil are studied. Particularly, characteristics of performance, features of temperature, pressure, steam chamber, distribution of residual oil in the later SAGD in extra heavy oil reservoir are deeply characterized. Combining these features and using physical simulation method, the feasibility of ISC in later SAGD in extra heavy oil reservoir has been discussed in terms of the effects of the oxidation kinetics characteristics , the thermal connectivity, the fuel supply, the coke deposit and the combustion front shape of extra heavy oil. Furthermore, the time when or before the steam chamber spreads to the edge of the SAGD well pair pattern is determined to be the optimum time to turn to ISC for the typical reservoir. By adding vertical wells for air injection in the middle of SAGD well pairs is the appropriate well pattern for ISC in the later SAGD. And perforating in the middle and lower interval is demonstrated to be the better method to control injection and production. Four stages in the process of ISC performance are determined and dissected. The study results indicate that stable combustion front shape and high production rate can be achieved after turning to ISC. Another 50.7% of the OOIP can be obtained in the ISC stage, regardless of 30% oil recovery in the SAGD stage.
For stress-sensitive heavy oil reservoirs, geomechanical responses of the reservoir are taken into account as they play an important role in the accurate simulation of all thermal recovery techniques, such as SAGD, or steamflood. However full-field numerical simulations of multi-physics processes by any coupling strategies are technically impossible with current computer CPUs. Under these conditions, analytical methods can be used as approximate techniques instead of numerical simulators, as they are much faster and yet are useful tools for preliminary forecasting and sensitivity studies. In analytical models, inclusions of all flow variables impacts into geomechanics frameworks make the equations so complex and almost impossible to solve. This paper provides a flow-based domain decomposition workflow for performing different analytical coupling schemes in different reservoir compartments.
Since the intensity and complexity of reservoir geomechanics vary over reservoir domain, one can divide the reservoir to some sub-domains and assess different geomechanical responses separately in each sub-domain.The presented analytical proxy, suggest decomposition of the whole domain in into two parts of “heated zone” and “wetted zone”, for rapid assessment of geomechanics. The heat flow equation was combined with mass and momentum convective transport equations to obtain an exact approach that correlates the saturation front of injected hot water to temperature front. The frontal velocities are dynamic interfaces for compartmentalization of the domain. In the heated zone , the total induced stresses, were considered due to both temperature and pressure increase, and in the wetted (saturated) zone beyond the temperature front , at each instance the total stress induced is only a function of pressure increase, and accordingly stress and strain induced are due to isotropic unloading. This technique provides a rapid estimate of geomechanical responses (stress and strain profile) in each part of the reservoir (near field and far field).
A numerical model was built and implemented in CMG-STARS for steam-flood case to show the robustness and applicability range of the model. The results were analyzed for synthetic case single-domain model and the model sensitivity on some reservoir parameters were checked, and at the same time geomechanical responses were not neglected anywhere (near-filed and far field) in the reservoir.
Techniques have been developed to experimentally and theoretically determine phase behaviour and viscosity reduction of CO2-heavy oil systems at high pressures and elevated temperatures. Experimentally, vapour-liquid phase boundaries (i.e., saturation pressure lines) and the swelling factors are measured by conducting PVT tests at pressures up to 11094.0 kPa and temperatures up to 362.75 K, respectively. The viscosity of CO2-saturated heavy oil is measured at 319.15 K. Theoretically, the heavy oil sample is respectively characterized as a single- and multi-pseudocomponent(s). An exponential distribution function is used to split the plus fraction of heavy oil up to C105+, while a logarithm-type lumping method is used to group the single carbon numbers (SCNs) into multiple pseudocomponents. Then, the Peng-Robinson equation of state (PR EOS) coupled with the modified alpha function is applied to quantify the phase and volumetric behaviour of the CO2-heavy oil systems. The binary interaction parameters (BIPs) for CO2-pseudocomponent(s) pair are tuned to match the measured saturation pressures. Compared with the characterization scheme of treating heavy oil as a single pseudocomponent, the absolute average relative deviation (AARD) for the predicted saturation pressures can be improved from 5.27% to 4.56% by characterizing the heavy oil as six pseudocomponents. With the optimum BIPs, the swelling factors are predicted by the PR EOS with and without the volume translation technique, respectively. It is found that the introduction of the volume shift to each (pseudo)component in the PR EOS is able to provide more accurate prediction in both characterization schemes with AARD of 1.88% (oil as a single pseudocomponent) and 1.39% (oil as six pseudocomponents), respectively.
Cyclic Solvent Injection (CSI) has been proposed as a follow-up process in post-CHOPS reservoirs. The oil recovery factor, which could reach over 50% in lab-scale, is mainly attributed to solution gas drive and foamy oil flow in CSI process. It has been suggested that pressure decline rate has impact on the behavior of solution gas drive and foamy oil flow in cold heavy oil production. However, the role of pressure decline rate in the post-CHOPS CSI process production has not been studied adequately. This work was intended to evaluate the effects of pressure decline rate in CSI process under post-CHOPS reservoirs’ conditions.
In this study, different pressure decline rate tests were conducted in a large cylindrical sand-pack model with a length of 30.48 m and a diameter of 15.24 cm. Single well was applied and connected with a mimic wormhole. In terms of oil recovery factor, the average recovery factor of each cycle increases with the increasing pressure decline rate. But considering the running time, the tests with smaller pressure decline rates showed better total recoveries compared with the tests with larger pressure decline rates. And the residual oil saturation pictures showed that the oil far away from the wellbore could be more easily recovered when a smaller pressure decline rate is applied. In terms of oil production rate, the CSI process production can be typically divided into two phases. In Phase 1, the production rate increases and reaches to the maximum value. In Phase 2, the production rate significantly declines. It was found that the test with a larger pressure decline rate had the higher production rate in Phase 1, while the test with the smaller pressure decline rate had the longest production time in Phase 1. The production rate hardly has dependence on the pressure decline rate in Phase 2. This indicated that for an optimized CSI process the pressure decline rate should be dynamically adjusted in order to have the best performance. The results also suggested that the minimum production pressure exists, below which no oil or marginal amount of oil was produced.
Co-injection of solvent with steam in SAGD has shown promise for enhancing oil rates as well as in reduction of energy and water consumption. Modeling and optimization of hybrid steam-solvent recovery processes with commercial numerical simulators can be very time consuming. In addition, the complex interaction of heat and solvent effects in mobilizing heavy oil at the vapour chamber boundary are often difficult to ascertain from the numerical models. Semi-analytical mathematical models can provide insight into the physics of the processes and may be used to estimate production rates and thermal efficiency in much less time.
In this study, an unsteady-state semi-analytical model was developed to predict the oil flow rate in the steam-solvent assisted recovery process. The model assumes a curved interface with transient temperature and solvent distribution in the mobile zone. It also accounts for transverse dispersion and concentration-dependent molecular diffusion for solvent distribution. The oil flow rate and interface profile are predicted at each time in an iterative fashion. The results show that the coefficient of diffusion-concentration function significantly affects the solvent penetration depth and its distribution. The semi-analytical model was able to predict oil production rates using different solvents co-injected with steam, in agreement with reported experimental data.
The proposed model reveals the complex interaction of heat and solvent solubility and diffusion as they affect mobilization and production of viscous oil. This model may be used to find the optimal operation parameters for the process over a range of different reservoir qualities and pressures, in a very time-efficient manner. The final outcome may lead to an efficient design of a steam-solvent recovery process that utilizes less water and reduces the amount of energy and gas emissions per barrel of oil produced.
Storage sites associated with depleted oil and gas reservoirs may contain many abandoned wellbores in addition to potentially unidentified wellbores. These wellbores have historically variable quality and quantity of cement that will have undergone ranging degrees of degradation. Wellbore performance in a single wellbore is dependent on the wellbore events (i.e. pressure and temperature changes) that occur within the life of the wellbore (Fourmaintraux et al., 2005; Gray et al., 2007).
There is significant uncertainty surrounding the integrity of existing wellbores due to a lack of data. Successful implementation of carbon capture and storage (CCS) will depend on solving the small-scale leakage problem associated with localized flow along wellbores. Our knowledge of oil wellbore performance under different life stages of a well is still weak. Consequently, each wellbore is unique and general conclusions about well integrity are difficult to ascertain from analyzing only a single well. Each wellbore is considered as a risk and robust tools are needed to allow for the assessment of the performance for wellbores and investigate wellbore leakage mechanism.
In this paper, a full lifecycle methodology is proposed to assess wellbore integrity as a measure of the risk of leakage. The methodology identifies the key elements to model the wellbore element and incorporates the use of a statistical approach to better understand the uncertainty in the risk estimation and interaction between various parameters controlling the model.