Al-Murayri, Mohammed Taha (Kuwait Oil Company) | Hassan, Abrahim Abdelgadir (Kuwait Oil Company) | Kamal, Dawood Sulaiman (Kuwait Oil Company) | Batôt, Guillaume (IFP Energies nouvelles, The EOR Alliance) | Cuenca, Amandine (Solvay, The EOR Alliance) | Butron, Jessica (Perm. Inc.) | Kantzas, Apostolos (Perm. Inc.) | Suzanne, Guillaume (Beicip Franlab, The EOR Alliance)
Foam has been extensively investigated as a method to improve the mobility control of non-condensable gases in the EOR context. Recently, there has been renewed interest in foam applied to steam injections. However, steam is a condensable gas and thus steam-foam requires special analyses that differ from classical foam assessments. This work presents the coreflood results of a steam-foam process evaluation for the Ratqa Lower Fars (RQLF) heavy oil reservoir in Kuwait.
Using specifically designed foaming surfactants, coreflood tests in the absence and presence of heavy crude oil are performed in native sandpack cores under RQLF reservoir conditions (220°C; 360 psi). In order to limit steam condensation due to the build-up of the foam pressure, steam has been supplemented with a small amount of non-condensable gas (nitrogen, about 1 - 5 mol.%). Interstitial velocity was decreased from 40 ft/day down to 1 ft/day (CWE). Phase equilibria at the core inlet were estimated based on thermodynamics flash calculations. From these calculations inlet steam quality was varied from 10 to 70 wt.%. In absence of oil, the apparent viscosity of the generated steam-foam is measured between 25 and 50 cP, depending on the interstitial velocity and inlet steam quality. Indeed, beside the classical shear-thickening behaviour observed with the decreasing flow rates, the critical or optimal steam quality is found to be closed to 30 wt.%. Furthermore, even at higher steam quality the foam is still stable and efficient with a viscosity higher than 25 cP. Experiments in the presence of crude oil were carried out under the same conditions in native cores at a steam residual oil saturation of 7% and 13%. These experiments showed that the optimal steam quality is shifted to approximatively 10 wt.%. Furthermore, the foam flow curve shows a shear-thinning behavior that is elaborated upon. Finally, the viscosity in the presence of heavy crude oil of the generated steam-foam is within the range of 7 to 22 cP, depending on the oil saturation and on the injection conditions. Considering the oil viscosity (2 to 3 cP) under the same conditions, this means that the foam effect should translate into efficient improved conformance control of the steam within the reservoir.
For the first time, an efficient and stable steam-foam is generated in coreflood experiments. The generated foam achieved high apparent viscosities, even in the presence of oil, and this has not been reported in the literature to date. The results presented here are far more than a proof of concept as they bring new evidences regarding steam-foam efficiency and mechanisms with heavy crude oil.
Real-time decision making, field surveillance, and production optimization improve the performance of existing operations to increase hydrocarbon recovery and reduce emissions. In this regard, the oil and condensate flow metering in offshore gas condensate platforms is always confronted by environmental, economic, and operational challenges resulting in uncertain production management plans. Although production forecasting of unconventional gas condensate systems is more challenging than for conventional wells, it is of great interest to support decisions by knowing the future of the wells as far as possible. The virtual flow metering techniques make it possible to utilize daily production data sets and extract information on how wells and reservoir will respond to different operational conditions. The objective of this study is to embed artificial intelligence algorithms in reservoir uncertainty modeling and present a mechanistically-supported data-driven model applicable for production forecasting of gas condensate wells with higher confidence. The outcome entails a new set of mathematical models, implemented using Apache Spark cluster computing engine with APIs in Python, that enables rigorous and robust optimization of the recovery process, designing and discovering hidden patterns in production data, and extracting reservoir information indirectly in seconds. The observations used to demonstrate the performance of the proposed hybrid model include 1600 well-testing data points together with 420 days of production history of an offshore gas condensate platform. The daily platform production is allocated efficiently to individual wells using a multilayer perceptron neural network model adaptively trained with well-testing and daily production datasets, and supported by the Energy and mass balance equations.
Thermal and solvent-based EOR methods are not applicable in many of thin post-CHOPS heavy oil reservoirs in Western Canada. Alkaline-surfactant flooding has been suggested as an alternative to develop these reservoirs. The main mechanism behind these processes has been attributed to emulsion-assisted conformance control due to the effect of synthetic and/or natural surfactants. Because nanoparticles (NPs) offer some advantages in emulsion stabilization, here we combine surface-modified silica NPs and anionic surfactants to enhance the efficiency of heavy oil chemical floods.
Based on the results of bulk fluid screening experiments, in the absence of surface-modified silica NP surfactant concentration should be tuned at the CMC (between 1 and 1.5 wt. %) to achieve the highest amount of emulsion. These emulsions are much less viscous than the originating heavy oil. However, at surfactant concentrations far from the CMC, complete phase separation occurs 24 hours after preparation. In the presence of surface-modified silica NP this emulsification was achieved at much lower surfactant concentration. The mixture of 0.1 wt. % anionic surfactant and 2 wt. % surface-modified silica NP produce a homogeneous emulsion of heavy oil in an aqueous phase. This observation was not observed when aqueous phase contains only either 0.1 wt. % anionic surfactant or 2 wt. % silica NP.
Preliminary tertiary chemical floods with water containing 0.1 wt. % surfactant and 2 wt. % surface-modified silica NP yielded an incremental oil recovery of 48 % OOIP, which is remarkably higher than that of either surfactant or NP floods with incremental recoveries of 16 and 36 % OOIP, respectively. Tertiary recovery efficiency, defined as ratio of incremental recovery factor to maximum pressure gradient during the tertiary flood, is six times greater for the surfactant/NP mixture than for the surfactant-only flood. This enhancement in recovery efficiency is of great interest for field applications where high EOR and large injectivity are desired.
Chemical flooding has been suggested as an efficient conformance control technique to develop many of thin post-CHOPS heavy oil reservoirs in Western Canada. In-situ formation of oil in water emulsions due to the effect of surfactant/natural soap has been reported as the main mechanism behind chemical EOR. In this work, the effect of surface-modified silica NPs to enhance the efficiency of surfactant to emulsify heavy oil (14,850 mPa.s and 980 kg/m3 at 25 °C, from the Luseland field) in water has been investigated.
Bulk fluid screening experiments were conducted using different surfactants and surface-modified silica NPs for selecting the best heavy oil emulsifier. Complementary experiments such as interfacial/surface tension, NP zeta potential and size measurements, and elemental analysis were conducted to understand the interactions between NPs and surfactant molecules.
In the absence of NPs, concentration of both anionic and cationic surfactants should be tuned within a narrow window, near CMC, to create stable heavy oil in water emulsions. It was found that there is a threshold for IFT, obtained at the CMC, which should be met to have stable oil in water emulsions. The created oil in water emulsions break easily at surfactant concentrations higher than the CMC, yielding IFTs higher than the threshold. This observation was also seen in a system containing dodecane. At the CMC of both anionic and cationic surfactants, the IFT between dodecane and an aqueous phase is negative, producing stable dodecane in water emulsions for over three months.
In the presence of surface-modified silica NPs heavy oil emulsification is achieved at surfactant concentrations much lower than the CMC. In this case, IFT is remarkably (54 %) reduced, well below the threshold value, due to the combined effect of 2 wt. % negatively-charged silica NPs and only 0.1 wt. % anionic surfactant. These results suggest that the repulsive interaction between negatively-charged NPs and anionic surfactant may result in pushing the surfactant molecules back towards the oil-water interface to enhance IFT reduction.
Most of the Nuclear Magnetic Resonance (NMR) log based permeability models require the estimation of the irreducible water saturation (Swirr). Several methods are available for calculating this parameter using NMR relaxometry. The most straightforward method with the lowest accuracy is to consider a fixed relaxation time (T2) value. It has been suggested to use a T2-cutoff equal to 10 ms for tight reservoirs. Another traditional experimental method involves centrifuging core plugs to Swirr. In this paper, an additional approach to separate free and bound water using NMR relaxation time is introduced. This method involves the area under the amplitude-T2 relaxation time graph.
A series of experiments were conducted on 81 core plugs. These samples are mainly from the Western Canadian Sedimentary Basin. Core plugs are from Montney, Nordegg, Mist Mountain, Red Beds, Doig, Killam, Lathom, York River, Wapiti, Teslimkoy, Kesan, and Ordivician Quartz formations. NMR measurements were obtained initially on the dry cores to establish the presence of any liquids that were not cleaned or any isolated porosity. The air permeability was measured using an in-house permeameter. The cores were then brine saturated in two steps of spontaneous imbibition followed by forced imbibition under vacuum. The Archimedes principle was used to measure the sample pore volumes. Porosity was subsequently calculated. NMR relaxation data were then acquired on the brine-saturated cores. Then the core plugs were centrifuged under air to an expected irreducible saturation. NMR relaxation times were obtained on all cores at Swirr.
NMR porosity, T2gm, Irreducible Bulk Volume (BVI), and Free Fluid Index (FFI) were calculated. Swirr was calculated with the three aforementioned methods. Excel Visual Basic for Applications (VBA) programming language was employed for analyzing the relaxation times. The Timur-Coates model was applied for permeability calculation using all the aforementioned Swirr estimation methods. Data were analyzed, and discrepancy analysis was conducted.
The implemented area analysis method has been used previously in reservoir typing based on formation types and also as a factor in one permeability model. However, this is the first time this approach is used in calculating FFI/BVI exclusively. This method is faster than conventional estimators, and it is the only method that can implement Timur-Coates based permeability models for logging tools. From the experimental point of view, only a single NMR measurement is needed. Centrifuging the cores is not necessary. The possibility of cracking these cores due to spinning is eliminated. This new approach is less computationally demanding, and calculations are easier to perform. It is proven that the fast peak area method is more accurate than the fixed T2-cuttoff and in some cases the centrifuge method.
A semi-analytical formulation of the spontaneous capillary imbibition is used to analyze the liquid intake of six shale samples, linking imbibition capacity and rate to lab-scale measurements. Moreover, a data-driven approach is utilized to examine the effect of mineralogy and porosity on the macroscale wettability of shales. According to the results, the presence of connected organic sub-layers lowers the destructive impacts of spontaneous water imbibition on hydrocarbon permeability. Furthermore, the intrinsic permeability, tortuosity, wettability, and initial and residual saturations are among the most influential factors influencing the water uptake during shut-in periods after hydraulic fracturing operations.
The fluid flow and transport phenomena are simulated through three-dimensional consolidated and unconsolidated digital rock slabs representing the geological rock types of the McMurray formation. Finite element simulations are carried out in order to solve the mixing advection-diffusion problem, and the effluent history is processed to predict the longitudinal dispersion coefficient and evaluate the efficiency of miscible displacement processes in the presence of microheterogeneities. According to the results, consolidated samples with non-uniform cementation show a different dispersion behavior when compared to both unconsolidated and uniformly cemented rocks, suggesting that the cementation pattern remarkably influences the miscible-flood performance.
Taking advantage of the analogy between hydraulic and electrical flows to facilitate the prediction of porous media characteristics is a longstanding practice in petroleum engineering. The relationship between hydraulic and electrical properties is widely used in well-logs interpretation and characterization of transport properties of porous media relying on the strong correlation between electric and hydraulic flow conductance. However, due to the lack of direct investigations, the similarity among their pathways/tortuosities is still unclear. It is a challenging and almost impractical task to identify the streamlines experimentally. Here a series of direct finite element numerical simulations are conducted within pore-level microstructures to extract and compare the streamlines of both electric and fluid flow currents and examine the accuracy of the analogy by predicting the petrophysical characteristics of the case studies. The fluid flow and electric transports are simulated through pore-level digital rocks of synthetic unconsolidated sand packs representing the Athabasca oil sands deposit as the second largest oil reserve in the world. The formation factor, porosity, and absolute permeability of the media under consideration are predicted, and consequently, the streamlines of both electric and hydraulic currents are extracted and compared in terms of length, shape, and pathways. According to the results, the fluid flow pathways pose differently and are longer than the homogeneous electric current streamlines. The ratio between the hydraulic and electric tortuosities follows a polynomial trendline, and a local extremum occurs at the porosity of eighteen percent. Pedotransfer functions for tortuosities, dimensionless permeability, and formation factor are proposed underpinning the rigorous relationships between transport processes in rocks.
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.
Mohammadmoradi, Peyman (University of Calgary) | Bashtani, Farzad (University of Calgary) | Goudarzi, Banafsheh (University of Calgary) | Taheri, Saeed (University of Calgary) | Kantzas, Apostolos (University of Calgary)
Due to the computational simplicity and time efficiency, pore network and morphological techniques are practical approaches for characterization of pore-scale structures. The methods are quasi-static and exploit pore space spatial statistics during invasion processes. Here, both procedures are evaluated applying the workflows to pore-level micro- and sub micro-scale images of Sandstone, Carbonate and Shale formations. A statistical approach is also utilized to improve the accuracy of Shale characteristics by spatial restoration of fragmentary parts of organic matter. Post-processing predictions include relative permeability and capillary pressure curves, absolute permeability, formation factor, and thermal connectivity. According to the results, the accuracy of pore network modeling in characterization of micro-CT images is compromised by the presence of limited number of network elements, ignoring the resistance of pore elements, multi-scale structures, and tight/weak connections represented by limited voxels. Pore network extraction affects the accuracy of petrophysical predictions and fluid occupancy profiles and also ignores the thermal and electrical properties of solid structure, including calcite, kerogen, quartz, etc. The pore morphological approach easily deals with a variety of rock configurations and resolutions and preserves connectivity and details of original images having more geometrical features than the pore network modeling. However, it predicts limited step-wised data points and realizations sourcing from its voxel-based nature. In addition, direct simulations confirm that stochastic conditional reconstruction of organic matter inside shale sub-volumes remarkably boosts the pore space connectivity and affects its predicted hydraulic properties.