Luo, Haishan (The University of Texas at Austin) | Li, Zhitao (Ultimate EOR Services LLC) | Tagavifar, Mohsen (The University of Texas at Austin) | Lashgari, HamidReza (The University of Texas at Austin) | Zhao, Bochao (The University of Texas at Austin) | Delshad, Mojdeh (The University of Texas at Austin) | Pope, Gary A. (The University of Texas at Austin) | Mohanty, Kishore K. (The University of Texas at Austin)
Polymer flooding has been commercially applied to a number of viscous oil fields in the past decade and gradually gained more popularity. Due to limited injectivity in viscous-oil reservoirs, a relatively low polymer viscosity is usually used to avoid excessive injection pressure. In such a case, mobility ratio of polymer solution to oil is much greater than one, which implies unstable flow and strong viscous fingering. Existing reservoir simulators lack the capability of modeling such a physical phenomenon. Since many viscous-oil reservoirs have high permeability contrast between layers, we are motivated to study, for the first time, the impact of crossflow between different layers considering the presence of viscous fingering.
Numerical modeling polymer floods with crossflow in a layered viscous-oil reservoir is difficult due to two major challenges: first is how to correctly allocate flow rates from the wellbore to multiple layers; and second is how to capture the viscous fingering effect without using excessively fine grids. To address the first issue, we developed an implicit well-rate-allocation model based on the potential method, which fully couples all the wellbore segments of each well with reservoir gridblocks to ensure a physical wellbore pressure. To deal with the second challenge, we implemented the effective fingering model, which is an upscaling model that lumps all the viscous fingers in a coarse grid block into one fictitious finger to allow for accurate estimation of fingering strength and growth during unstable flows. Both models were validated individually against the analytical solution or experimental data.
The integrative module including the two new capabilities was used to simulate a polymer flood following a waterflood in a layer-cake reservoir in North America with moderate oil viscosity. We observed the fast propagation of water fronts and small fingering fraction in high permeability layers during the waterflooding phase, indicating active channeling and viscous fingering. The subsequent polymer flooding minimized both factors of oil bypassing and led to stable flow and high sweep efficiency. Without the implicit well-rate-allocation model, crossflow was overestimated and wellbore pressures of different well blocks were not consistent. Without the effective-fingering model, oil recovery was overestimated due to the lack of accounting for viscous fingering. The simulation results indicated that polymer flooding improved both displacement and sweep efficiencies.
The model has shown comprehensive capabilities in reservoir simulations of polymer floods including unstable floods and crossflows between layers. This is a major significance to optimization of non-thermal viscous-oil EOR projects and also making more informed operational decisions for field developments.
This paper describes the key learnings from detailed data analysis and the application of a probabilistic workflow to achieve a validated dynamic simulation model of the large scale pilot (LSP) that was successfully completed in the 1st Eocene reservoir in the Wafra field, PZ. The development of the validated dynamic simulation model was a key success measure identified for the pilot to assist in the broader Wafra full field steamflood development decision making process.
The pilot consisted of sixteen 2.5 acre inverted 5 spot patterns (16 injectors and 25 producers) with associated steamflood and production facilities. In the execution of this work, detailed analysis of production, pressure, temperature, oil viscosity and geochemical data was conducted to develop key insights into the pilot behavior.
A 1st Eocene LSP steamflood model was successfully validated with production and injection data through the primary and steamflood periods. The validation method took into account uncertainties in static and dynamic properties. A probabilistic methodology using design of experiments on uncertainty parameters, combined with error-filters on multiple responses of rates, pressure and temperature was used in the study. The confidence in the final validated model was high with excellent agreement achieved between pilot and model data. The validated model also captured regional behaviors observed in the production data analysis. The validated model captured both pre-steam and post-steam behavior in the pilot. The successful execution of this work has resulted in an estimate of carbonate steamflood recovery possible in the 1st Eocene reservoir in Wafra.
Steam assisted gravity drainage (SAGD), as a commercially proven high ultimate recovery process for heavy oil and bitumen, is energy intensive and the process is estimated to be economic when oil prices are above certain level. Under conditions were oil price is below the economic threshold of the project, steam injection can be reduced and or completely stopped for a period of time; and can resume thereafter when conditions change. Although the process performance and efficiency may get impaired compared to the continuous steam injection case, it is still essential to evaluate and optimize the operating strategy after steam injection is re-initiated.
This study, thus, aims at providing insights about impact of steam injection interruptions on SAGD project performance. Optimization strategy of SAGD process in cases, where steam injection is interrupted, is also discussed. Field scale simulation of the process has been conducted with actual reservoir models of different geological formations to investigate the optimum operating strategy. An economic model is employed to evaluate the impact of operating strategy on the net present value (NPV) of the project. The parameters such as initial steam injection period, shut-in period, steam injection rate during the re-initiation period etc. are optimized for Athabasca type oil sand reservoirs.
The results indicate that there are several key mechanisms in the life cycle of a SAGD process that must be considered in the simulation model to reflect the field scale behavior and obtain the optimum operating strategy; otherwise, due to the mechanistic simplicity of the models, the results might be at best semi-quantitative and directional. Among all the mechanisms, the impact of temperature on the basic petrophysical properties of reservoir rocks, such as relative permeability, has been found to play an important role in the oil recovery and significantly affects the optimization results. Despite the complex behavior of the SAGD process, when the process is interrupted, there would be an optimum shut-in period in which the oil recovery can be maximized. The optimum length of interruption depends on the initial period of steam injection.
In the petroleum industry, accurately estimating wellbore heat loss in thermal recovery processes remains a critical problem. One difficulty lies in simulating heat transfer and fluid dynamics within wellbore annuli. A literature survey shows that the state-of-the-art thermal wellbore simulators use empirical correlations to calculate the heat loss through wellbore annuli. As more sophisticated wells have been drilled, there is a growing need for a more detailed wellbore annulus heat transfer model. In this study, a 2D transient mathematical model is proposed for the conjugate natural convection and radiation within wellbore annuli. The governing equations consist of a continuity equation, a vorticity transfer equation, an energy transport equation and a radiative transfer equation (RTE). A finite volume approach with a second-order upwinding scheme is implemented for discretization. Newton-Raphson iterations are deployed for linearization. The algorithm is validated by consistence in simulation results compared with benchmark numerical solutions with the Rayleigh number up to 107. Parameters such as an aspect ratio, a radius ratio and a conduction-to-radiation coefficient are examined. A case study on vacuum insulated tubing heat transfer using Marlin Well A-6 data demonstrates the merits of the developed program by the consistence of simulation results compared with field measurements.
A new method is proposed for simple and reliable viscosity calculations for mixtures of heavy oils or bitumens with light solvents. This method has three characteristics that give it an advantage over current approaches. First, it is simple to implement. Second, it faithfully portrays the viscosity of mixtures of Newtonian fluids up to at least 190°C. Third, and most importantly, it makes use of the recent discovery that, for any given heavy oil, all light solvents can be represented at low concentrations by the same mole-fraction-weighted form of the Arrhenius equation. For high concentrations, an additional, empirical power-law term, whose exponent was determined from new viscosity data for toluene–heavy oil mixtures, was developed. The average absolute relative deviation (AARD) between 307 experimental viscosity measurements and the new equation was 20.1%. The data set included mixture viscosities from 1 to 440,000 mPa·s, at temperatures from 12 to 190°C. The largest disagreements were -63% and 101% of the measured values. It is suggested that a substantial portion of the observed disagreements arose from combined errors in the experimental measurement of viscosities, solvent compositions, and average molecular weights of the oils.
It was found that the one key constant for describing concentration effects could be correlated reasonably well with the viscosity of the dead oil. When this constant was calculated from the resulting correlation, the AARD for the initial data set increased by only a modest amount to 27.7%. The disagreements were somewhat larger when the new method was used to make pure predictions of two published data sets. Accordingly, it was concluded that the new approach can be used to make reasonable predictions (possibly within a factor of 2) of solvent–oil mixture viscosities simply from accurate knowledge of the viscosities and molecular weights of the dead oil and solvent. However, substantially better predictions may be obtained for any light solvent if the above information is supplemented with even relatively inexpensive laboratory data taken with a lower-volatility solvent such as toluene.
The Saleski steam injection Pilot in Alberta, Canada, has proven recovery from a naturally fractured bitumen reservoir in the Grosmont carbonate. Non-condensable gases (NCG) generation and their behavior and flow in the reservoir have a significant effect on the performance of the thermal steam injection processes. This becomes even more important in a carbonate fractured reservoir where gas saturation and distribution largely impact Gas Oil Gravity Drainage (GOGD) process. Therefore, NCGs were likely instrumental in the successful recovery observed in the Grosmont Pilot. Understanding the role played by NCGs in the Cyclic SAGD process at Saleski can lead to insights and improvement in the recovery process.
The Saleski Pilot, a joint venture between Laricina Energy Ltd. and Osum Oil Sands Corp. started in 2010 to exploit the Grosmont carbonate fractured reservoir. Saleski pilot operation was ended in September 2015 as the goals of the project were achieved. During close to five years of operation of the pilot a great deal of data on all important parameters were collected. Gas production numbers as wells as gas composition were also measured and Laricina Energy Ltd. has agreed to make it available to the public. The NCG production data is reviewed for the whole period of operation and total GOR as well as individual gases GOR are calculated. This information is used to calibrate a kinetic reaction model which is implied in the numerical simulation work in order to examine the effect of NCG.
Results from lab experiment and Saleski Pilot show gas production higher than initial solution gas of the bitumen with substantially higher than expected concentration of CO2. To further understand these observations, a kinetic model is developed and calibrated to the observations. This model was then implemented into reservoir numerical simulations to develop further insights into thermal bitumen recovery from the Grosmont and improve history matching practice of the Saleski pilot.
Gas generation, behavior, and flow inside the reservoir during a thermal steam process in a carbonate fractured reservoir have not been investigated previously. This work presents invaluable field data on gas production in Grosmont reservoir and further assesses the possible effect of NCG presence on recovery mechanisms.
Precipitation and deposition of asphaltene in reservoir rock can cause formation damage and reduce the fluid mobility, resulting in significant loss of the hydrocarbon production. A detailed study on asphaltene deposition behavior in porous media improves the understanding of asphaltene-induced formation damage and provides the possible solutions for preventing and/or controlling formation damage. Carbonate reservoirs in Western Canadian Sedimentary Basin (WCSB) have several challenges for enhanced oil recovery process (i.e. heterogeneity, the high viscosity of oil). Asphaltene-induced formation damage makes it more difficult for any process to recover heavy oil from such complex reservoirs.
For advancements in understanding asphaltene deposition in fractured carbonate formations, asphaltene deposition behavior was analyzed with the help of pore scale observation during hydrocarbon solvent injection. Three types of solvent (nC5, nC7, and nC12) were used for solvent injection for extra-heavy oil recovery (30,000 cp at 22 °C). A uniquely designed heterogeneous-fractured micromodel imitating the fractured carbonate reservoirs was used for pore scale observation. A high-quality camera along with a microscope was utilized to capture images. SEM analysis was also performed on precipitated asphaltene to visualize the difference in the structure of the asphaltenes precipitated with different solvents.
Observation through this study revealed that besides the amount of asphaltene deposition, the distribution pattern of asphaltene deposition could also be different when using various types of solvent. By increasing carbon number of solvent from C5 to C12, the amount of precipitation decreased while the wider distribution of deposited asphaltene was observed in a fractured-heterogeneous porous media. Moreover, it was noted that asphaltene could be deposited in different shapes that may or may not block the pore throat. The dominant flow mechanism (either diffusion dominant or viscous dominant) will affect the shape of deposited asphaltene. Asphaltene can deposit perpendicular or parallel to the flow direction which is in the form of parallel rope-shape deposits. The perpendicular deposits (usually in dominant diffusion flow), mainly observed in small pore throat, can block the diffusion path. However, the parallel shape deposits (typically in viscous dominant flow) will not significantly impede the flow path.
The results of this study improve our understanding of different aspects of asphaltene-induced formation damage in the fractured carbonate reservoirs such as; susceptible locations of asphaltene deposition, different types/shapes of deposition which may or may not result in pore blocking, and effect of flow behavior and heterogeneity on asphaltene deposition and formation damage. The more we know about the asphaltene deposition in such heterogeneous reservoirs, the better we can control the formation damage and increase the effectiveness of heavy oil recovery processes.
One of the ways to improve the efficiency of steam injection is to use chemicals as an additive to alter interfacial properties. Historically, this has been tested using surfactants, which are expensive and thermally unstable. Therefore, commercial applications have been highly limited in that area over the past three decades. In conjunction with recent efforts using new generation materials as EOR agents, we performed a screening study to identify the potential chemicals/materials for heavy-oil recovery and to investigate the applicability of selected new generation chemicals as interfacial properties modifiers at steam temperature.
Different experimental methods, including capillary imbibition tests, i.e. contact angle and interfacial tension measurements, were combined to understand the mechanism of alteration surface interplay (wettability and interfacial tension) using different chemical agents. Capillary imbibition tests were conducted to study the potential of these chemicals to alter wettability and rock/chemical interactions on limestone and aged sandstone cores at high temperature. Pendant drop interfacial tension (IFT) and contact angle measurements were performed using a high pressure and high temperature cell under the same -steam- conditions.
Seven different chemical agents including a high pH solution (sodium metaborate), an ionic liquid, a cationic and an anionic surfactant, and nanofluids (aluminum and zirconium oxides) were tested in this study. Indiana limestone samples were saturated and aged in heavy oil with a viscosity of 6,000 cp. Capillary imbibition tests were conducted under high temperature (between 90°C and 180°C) and high pressure (185 psi) conditions using a newly-manufactured visual cell. The production rate and ultimate recovery were used to evaluate the capability of different chemicals changing the interfacial properties and their thermal stability at steam temperature. Contact angles between heavy oil and calcite plates were measured under the same conditions. Finally, the stability of the chemicals was measured through settlement tests at steam temperature conditions as well as TGA (thermal gravimetric analysis). The combination of all these results helped identify the applicability of the selected chemicals under steam conditions for carbonates. Technical and economic limitations for each chemical as well as the way the chemical contributes to recovery (wettability alteration or IFT reduction) were identified.
Investigation of interfacial properties alteration induced by new generation chemicals at high temperature is helpful in the selection and application of efficient and economical chemicals in steam based heavy-oil recovery methods.
Liu, Hao (China University of Petroleum-Beijing) | Cheng, Linsong (China University of Petroleum-Beijing) | Li, Chunlan (China University of Petroleum-Beijing) | Xiong, Hao (China University of Petroleum-Beijing) | Xiao, Peng (China University of Petroleum-Beijing)
In order to save energy and to be more environmentally friendly, Expanding-Solvent SAGD (ES-SAGD) is proposed by adding solvents into the injection vapor. However, much heat may still be wasted due to early steam condensation, which is associated with heat transfer and phase behavior within the steam chamber during ES-SAGD process. The objectives of this paper are to study temperature distribution within the steam chamber and to evaluate the overall heat loss rate of ES-SAGD by using a semi-analytical model.
In the mathematical model, temperature and mass profiles within the chamber are implicitly represented rather than explicitly treated as beyond the edge of the chamber. By considering early steam condensation and phase change within the chamber, the model dynamically couples heat transfer equation and mass transfer equation. This approach allows us to clearly observe the heat loss rate within the chamber as well as to the overburden. The model is solved iteratively after the steam velocity at the chamber edge is determined by introducing theory of steam front stability. Finally, the proposed model is validated by comparing calculated temperature profile with the numerical simulation results.
The results of the analytical study reveal that, for solvents with smaller molecular mass, the temperature starts dropping deeper from chamber edge to the injection end, and the temperature gap between the two ends is larger than that of heavier solvents. Based on the calculated temperature profiles, it is considered that heavier solvents may not improve thermal efficiency, while lighter solvents can help to reduce heat loss rate to the overburden especially in the late stage of ES-SAGD. Moreover, although lighter solvents are always used to relieve the heat losses to the overburden, the heat wasted inside the steam chamber is severer than that of heavier solvents. Therefore, considering the total heat losses to the overburden and inside the steam chamber, the steam/oil ratio (SOR) of ES-SAGD in some ultra-thick reservoirs may be unfavorable even compared with that of SAGD.
On the basis of analysis for heat loss rate to the overburden and inside the chamber, several guidelines are presented to select solvent for ES-SAGD process. The proposed guidelines will help to better predict and design the future ES-SAGD heavy oil recovery projects.
Huang, Shijun (China University of Petroleum-Beijing) | Liu, Hao (China University of Petroleum-Beijing) | Xue, Yongchao (China University of Petroleum-Beijing) | Xiao, Peng (China University of Petroleum-Beijing) | Xiong, Hao (China University of Petroleum-Beijing)
More than half of heavy oil reservoirs in Western Canada are less than 5m thick and SAGD is generally not thought to be economically viable for such kind of reservoirs due to lack of Potential Energy of Gravity and significant heat losses to the overburden. Solvent enhanced steam flooding (SESF), a kind of enhanced steam flooding by co-injecting solvent with steam, has shown to be promising in enhancing oil rates in thin reservoirs. In this paper, a semi-analytical model is established for predicting production performance of SESF.
The model is mainly built based on Energy Conservation and Fick's Law to predict the steam front position as well as the solvent concentration profile in the reservoir. Besides, the blocking effects of steam condensate on solvent diffusion is modeled by introducing the oil-water two phase flow theory. Then, the model is divided into three parts corresponding to the three production stages of SESF, and they are semi-analytically solved successively. The proposed model is validated by comparing calculated oil production rate with the results of a numerical simulation method. The results indicate that the enhancement of oil production rate mainly happened in the early stage of the process which is achieved by the combining effects of heat and convection-enhanced mixing of solvent and heavy oil. On the basis of a sensitivity analysis for performance of SESF, it is realized that the operating thickness and solubility of a solvent are proportional to steam oil ratio reduction of SESF compared to conventional steam flooding. Besides, a relatively lower injection rate and a longer well spacing may result in higher thermal efficiency increment due to longer contacting time of solvent with crude oil.
Piloting SESF in a field has many challenges, especially when considering its main economic factors: production increase and solvent cost. Therefore, it is anticipated that considering the dynamic mass transfer and blocking effect of accumulated condensate on solvent diffusion in SESF process, which are important mechanisms of SESF with inadequate understanding in literatures, the newly developed model will help to better predict and design the future SESF heavy oil recovery projects in thin pay zones.