Ole Torsæter

Metal oxide-based nanoparticles: revealing their potential to enhance oil recovery in different wettability systems

This paper presents systematic studies of hydrophilic metal oxide nanoparticles (NPs) dispersed in brine intended to reveal their potential to enhance oil recovery (EOR) in various rock wettability systems. The stability in suspension (nanofluid) of the NPs has been identified as a key factor related to their use as an EOR agent. Experimental techniques have been developed for nanofluid stability using three coupled methods: direct visual observation, surface conductivity and particle size measurements. The use of a dispersant has been investigated and has been shown to successfully improve metal oxide nanofluid stability as a function of its concentration. The dispersant alters the nanofluid properties, i.e. surface conductivity, pH and particle size distribution. A two-phase coreflood experiment was conducted by injecting the stable nanofluids as a tertiary process (nano-EOR) through core plugs with various wettabilities ranging from water-wet to oil-wet. The combination of metal oxide nanofluid and dispersant improved the oil recovery to a greater extent than either silica-based nanofluid or dispersant alone in all wettability systems. The contact angle, interfacial tension (IFT) and effluent were also measured. It was observed that metal oxide-based nanofluids altered the quartz plates to become more water-wet, and the results are consistent with those of the coreflood experiment. The particle adsorption during the transport process was identified from effluent analysis. The presence of NPs and dispersant reduced the IFT, but its reduction is sufficient to yield significant additional oil recovery. Hence, wettability alteration plays a dominant role in the oil displacement mechanism using nano-EOR.

Stochastic Reconstruction of Chalk from 2D Images

We study the stochastic reconstruction of the microstructure of chalk, from limited morphological information that may be readily extracted from 2D images of the pore space. Backscatter Scanning Electron Microscope images of a North Sea chalk sample are analyzed to determine de-scriptors of pore space morphology, such as the void-phase autocorrelation function, and void- and solid-phase chord distribution functions. This information is used to constrain the stochastic reconstruction of the chalk sample in 2D and 3D by a simulated annealing method. Quantitative analysis of 2D reconstructions using different morphological constraints reveals that imposing chord distribution functions results only in minor improvement over what is achieved by using the void-phase autocorrelation function as the only constraint. This result is further verified by geometric and topological characterization of a 3D replica of the sample generated using only autocorrelation function constraints. Pore and throat size distributions determined by 3D pore space partitioning methods, are consistent with mercury porosimetry results. The predicted permeability and formation factor are shown to be in very good agreement with experimentally determined values.

Stochastic Reconstruction, 3D Characterization and Network Modeling of Chalk

Abstract Systematic studies involving stochastic reconstruction, geometric and topological characterization, and network modeling of chalk, aiming at computation of petrophysical properties, are reported. The numerical chalk models are constructed exclusively from limited morphological information obtained from 2D backscatter scanning electron microscope images of the microstructure. Two different stochastic reconstruction methods are considered: conditioning and truncation of Gaussian random fields (GRF), and simulated annealing (SA). The potential of initializing the simulated annealing reconstruction with input generated using the Gaussian random fields method is evaluated and found to accelerate significantly the rate of convergence of simulated annealing reconstruction. This finding is important because the main advantage of simulated annealing method, namely its ability to impose a variety of reconstruction constraints, is usually compromised by its very slow rate of convergence. A detailed description of the chalk microstructure in the form of 3D volume data is essential for the prediction of petrophysical properties from first principles. The prediction of absolute permeability and formation factor directly from such information are considered first. The prediction of absolute permeability, formation factor and mercury–air capillary pressure curves are then considered using approximate network models constrained by information (pore- and throat-size distributions, coordination number) obtained from geometric and topological analysis of the reconstructed pore networks. Such information is extracted from the 3D volume data using morphological skeletonization and pore space partitioning methods. Very good agreement between the predicted and measured data is found for samples of North Sea chalk. On the basis of this study, it is concluded that (a) stochastic reconstruction from limited morphological information reproduces the essential features of pore geometry and connectivity of chalk, and (b) network modeling techniques can be used to predict petrophysical properties of chalk based on geometric and topological information of the stochastically reconstructed media.

An experimental study of the relationship between wettability and oil production characteristics

Abstract Wettability tests have been performed using two different North Sea sandstones and three different fluid systems composed of a NaCl-brine and pure n -decane, or n -decane with additives. Oil displacement experiments by water injection with composite cores of the same types of sandstone, and the same fluid systems, have also been done. It has been shown that it is possible to modify the wettability characteristics of the two sandstones from water-wet to neutral-wet, and further to oil-wet, by addition of small amounts of organic acid or organic base to the oil. Wettability indices obtained by the Amott and USBM tests are consistent. Water injection into two composite sandstone cores with fluid systems giving the three different states of wettability mentioned above, responded as expected. The water-wet cores had the highest oil recoveries at water break through. The non-water-wet systems all exhibited a significant tail production of oil. The highest ultimate oil recoveries were obtained for the neutral-wet systems, and the lowest recoveries were given by the oil-wet systems. Change in wettability by addition of organic acids or bases to the oil is likely to be due to adsorption of the additive on the surface of the rock. The mechanism of wettability alteration is thus similar to what can be obtained by the addition of a water-soluble surfactant to the brine. The only difference is that an additional interphase mass transfer step is involved for the oil-soluble, practically water-insoluble, additive to reach the rock surface by diffusion.

The effect of surfactant concentration on the Amott wettability index and residual oil saturation

The objective of this study was to determine the influence of surfactant concentration, surfactant adsorption and interfacial tension between oil and aqueous phase on the Amott wettability index and the residual oil saturation during waterflooding. The flooding experiment and Amott wettability tests were performed on 60 core plugs at a temperature of 50°C, with n-heptan and 1.5 wt.% NaCl brine with 12 different concentrations of surfactant. The anionic surfactant used in the experiments was n-dodecyl-o-xylene-sulfonate with a well-known interfacial tension vs. concentration relationship. The core material was Berea Sandstone with an average porosity of 19% and an average permeability of 104 md. Wettability indexes were determined by the standard Amott test involving spontaneous uptake of fluids and forced displacement steps. The Amott wettability index and residual oil saturation vs. surfactant concentration and an adsorption isotherm of the surfactant on Berea sandstone were the main quantities determined. The results show that increased surfactant concentration results in a change in wettability from strongly water wet to weakly oil wet. The residual oil saturation decreases with increasing surfactant concentration. A dramatic change in wettability from water wet to neutral and over to oil wet occurs in a narrow range of low concentrations which coincide with the critical micelle concentration. In this range the measured adsorption was only approximately 10% of the maximum value of the adsorption isotherm. Higher surfactant adsorption at higher concentrations did not have significant effect on the observed wettability.

Reconstruction of chalk pore networks from 2D backscatter electron micrographs using a simulated annealing technique

We report the stochastic reconstruction of chalk pore networks from limited morphological information that may be readily extracted from 2D backscatter electron (BSE) images of the pore space. The reconstruction technique employs a simulated annealing (SA) algorithm, which can be constrained by an arbitrary number of morphological descriptors. Backscatter electron images of a high-porosity North Sea chalk sample are analyzed and the morphological descriptors of the pore space are determined. The morphological descriptors considered are the void-phase two-point probability function and lineal path function computed with or without the application of periodic boundary conditions (PBC). 2D and 3D samples have been reconstructed with different combinations of the descriptors and the reconstructed pore networks have been analyzed quantitatively to evaluate the quality of reconstructions. The results demonstrate that simulated annealing technique may be used to reconstruct chalk pore networks with reasonable accuracy using the void-phase two-point probability function and/or void-phase lineal path function. Void-phase two-point probability function produces slightly better reconstruction than the void-phase lineal path function. Imposing void-phase lineal path function results in slight improvement over what is achieved by using the void-phase two-point probability function as the only constraint. Application of periodic boundary conditions appears to be not critically important when reasonably large samples are reconstructed.

Improved Oil Recovery by Hydrophilic Silica Nanoparticles Suspension: 2 Phase Flow Experimental Studies

In last decade, a number of papers about nanoparticles studies have been published related to its benefit for oil and gas industries. Some of them discussed about the potential of nanoparticles for enhanced oil recovery (EOR) in the laboratory scale. One of possible EOR mechanisms of nanofluids has been described as disjoining pressure gradient (Chengara, 2004, and Wasan, 2011). The benefit of using silica nanoparticles was explained by Miranda (2012). Hence, the present study objective is to investigate the potential of hydrophilic silica nanoparticles suspension as enhanced oil recovery agent and find out the main mechanisms of nanofluids for EOR. In this study, hydrophilic nanoparticles with average particle size of 7 nm were used in both visualization glass micromodel flooding experiments and core flooding experiments. A water-wet transparent glass micromodel and Berea sandstone cores with 300-400 mD permeability were used as porous medium. Synthetic brine was used as disperse fluid for nanoparticles. In order to investigate the recovery mechanisms of nanofluids, interfacial tension (IFT) and contact angle between different concentration nanofluids and crude oil have been measured by using spinning drop and pendent drop methods. The experimental results indicate that the nanofluids can reduce the IFT between water phase and oil phase and make the solid surface more water wet. In the visualization glass micromodel flooding experiments, it was observed that nanofluids can release oil drops trapped by capillary pressure, while the high concentration nanofluids stabilized oil-water emulsion. For the core flooding experiments, nanofluids can increase recovery about 4-5% compared to brine flooding. These results indicate that these nanoparticles are potential EOR agents. The future expectation is that nanoparticles could mobilize more oil in the pore network at field scale to improve oil recovery.

Wettability alterations due to an oil soluble additive

Abstract Wettability tests were performed using Berea sandstone, NaCl brine, and n-decane with different concentrations of hexadecylamine. The tests were performed on cores both with and without initial water. Adsorption tests of the additive have also been performed on crushed Berea sandstone. It was shown that it is possible to modify the wetting properties from water-wet to neutral by the addition of a small amount of hexadecylamine. This is most likely due to adsorption of the additive on the rock surface. The resulting wetting properties were observed to be dependent on the concentration of the additive, and the effect of the additive is higher for cores not having any initial water. It was also observed that both residual oil saturation and irreducible water saturation decrease when the cores approach neutral wetting properties. This could be due to the formation of two continuous phases through the cores and less capillary trapping. The adsorption results show an increasing adsorption with increasing concentration, and the amount of adsorption for dry sandstone is twice the amount of initially water-saturated sandstone.

Experimental PVT Property Analyses for Athabasca Bitumen

Heavy oil and tar sands are important hydrocarbon resources that are destined to play an increasingly important role in the oil supply of the world. A huge proportion of total world oil resources are in the form of these highly viscous fluids. The main recovery mechanism for these kinds of reservoirs is to somehow reduce their viscosity by the application of heat. In these extra heavy oil reservoirs, the reservoir has almost no injectivity, and therefore conventional steam flooding is hard to conduct. Steam Assisted Gravity Drainage (SAGD), however, reduces the viscosity of bitumen in place and the heated bitumen drains due to gravity forces towards the production well, where it is produced. Modeling and evaluating the production mechanisms in this process requires a thorough understanding of multi-phase flow parameters like relative permeability.Relative permeability data depend on a number of different parameters among others temperature and fluid viscosity. Viscosities of the flowing fluids drop with temperature, which can affect the relative permeability data. There has been a long debate on the actual impact of temperature on the relative permeabilities. Although some authors have reported saturation range shifts and relative permeability curve variations by temperature, others have attributed these variations to artifacts inherent in the methods used and the systems tested. Viscous instabilities and fingering issues have been blamed for temperature dependencies reported, and some researchers have reported that relative permeability data changes due to oil/water viscosity ratio changes at different temperatures.The variations in the experimental conditions have resulted in different and even contradictory results. There is specifically few experimental works conducted on Athabasca oil systems, and previously reported trends mainly apply to less viscous oils. This implies that the actual effect of temperature on flow behavior of fluids in the rock is case specific. Due to the contradictory reports and conclusions, which are due to variation in the systems being tested, it seemed necessary to conduct our own core flooding experiments, and investigate the curves of relative permeability. The objective was to obtain the imbibition relative permeability curves in an Athabasca oil type reservoir at different temperatures and oil viscosities, and figure out any possible trends of variations with temperature.Before conducting the core flooding experiments, some fluid behavior experiments were done to figure out the properties of bitumen used in this study. These include fluid compositions, density, viscosity, molecular weight and oil/steam interfacial tension. These properties were further used in numerical simulation studies.Core floodings were conducted on glass bead packs and sand packs saturated with heavy oils with varying viscosities. Displacement experiments with water were performed at different temperatures, and unsteady-state method of relative permeability measurement was conducted. The relative permeability data were determined by history matching the oil production data and pressure differential data in each experiment.Results indicated a change in the water saturation range in the oil-water relative permeability curves. The shift was towards higher water saturations, meaning an increase in irreducible water saturation and a decrease in residual oil saturation. Regarding the shape of relative permeability data, no unique trend of either rising or falling with temperature was found for oil and water relative permeability curves. The viscous instabilities are believed to be present in the experiments.As the same saturation range shift occurs by comparing the results at the same temperature level and by only changing the oil viscosity, this suggests that the temperature dependency of relative permeabilities can be attributed to the drop in oil to water viscosity ratio by temperature.The variations of relative permeability data with temperature was therefore found to be more related to artifacts in the experimental procedures like viscous fingering, and fluid viscosity changes than fundamental flow properties.Numerical simulations were accomplished on field scale SAGD and ES-SAGD (Expanding Solvent SAGD) operations testing the effect of relative permeability curves. Temperature dependent relative permeability data were tested and Oil production was found to be strongly dependant on the end point data. It is therefore suggested to use this option as a matching criterion when trying to history match SAGD field data.Since the main experimental part of this study deals with temperature dependency of relative permeability data, the introduction of this thesis is totally devoted to introducing this concept and its measurement methods and a literature review on the works performed so far. The main thesis is composed of three main parts, the fluid behavior experiments on bitumen, one-dimensional flow studies and multi-dimensional flow part. The results of fluid behavior experiments are given in chapter 2. Chapters 3 and 4 are devoted to one-dimensional flow works and chapters 5 and 6 present the part of this thesis dealing with two and three-dimensional flow. It should, however, be mentioned that chapters 4 to 6 can be read independently, as the contents of these chapters are taken from previously published papers with some minor revisions.

Multiphase Flow in Fractured Reservoirs

Naturally fractured reservoirs represent a complex class of reservoirs. Multiphase flow in such reservoirs adds to the complexity and has been studied extensively over the last few years.