Recovery Mechanisms and Relative Permeability ... - ACS Publications

Mar 21, 2013 - Recovery Mechanisms and Relative Permeability for Gas/Oil Systems ... Centre for Enhanced Oil Recovery and CO2 Solutions, Heriot-Watt U...
0 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/EF

Recovery Mechanisms and Relative Permeability for Gas/Oil Systems at Near-Miscible Conditions: Effects of Immobile Water Saturation, Wettability, Hysteresis, and Permeability S. Mobeen Fatemi* and Mehran Sohrabi Centre for Enhanced Oil Recovery and CO2 Solutions, Heriot-Watt University, Edinburgh EH14 4AS, United Kingdom ABSTRACT: Near-miscible gas injection represents a number of processes of great importance to reservoir engineers, including hydrocarbon gas injection and CO2 flood. Very little experimental data is available in the literature on displacements involving very low interfacial tension (IFT). In this paper, we present the results of a series of two- and three-phase gas injection (drainage) and oil injection (imbibition) core flood experiments for a gas/oil system at near-miscible (IFT = 0.04 mN m−1) conditions. Two different cores, a high-permeability (1000 mD) and a lower permeability (65 mD) core, were used in the experiments, and both water- and mixed-wet conditions were examined. The results show that, despite a very low gas/oil IFT, there is significant hysteresis between the imbibition and drainage oil and gas relative permeability (kr) curves in the 65 mD core. Hysteresis was less for the 1000 mD core (in comparison to the 65 mD core), but it still could not be ignored. Near-miscible kr hysteresis was significant for both water- and mixed-wet systems. The presence of immobile water in the water-wet cores improved oil relative permeabilities but reduced gas relative permeabilities in both imbibition and drainage directions. As a result, oil recovery for gas injection experiments improved when the rock contained immobile water. Both oil and gas relative permeabilities reduced when the rock wettability was altered to mixed-wet from water-wet, and as a result, oil recovery by gas injection in the mixed-wet rock was less than that obtained under water-wet conditions. We offer explanations for these observations based on our understanding of the pore-scale interactions and mechanisms, the distribution of fluid phases, and their spreading behavior. The results help us better understand the impact of some of the important parameters pertinent to kr and its hysteresis, especially in very low IFT gas/oil systems and mixed-wet rocks. Understanding these effects and behavior is important for the improved prediction of the performance of gas injection and water-alternating-gas (WAG) injection in oil reservoirs.

1. INTRODUCTION Proper core preparation for special core analysis (SCAL) can be difficult and time-consuming. Among different core preparation procedures, establishing immobile water saturation and restoring wettability of reservoir rocks are especially difficult. As a result, usually SCAL analyses are performed on cleaned core samples, which are without establishment immobile water saturation and/or water-wet. However, it is generally accepted that many oil reservoirs are mixed-wet.1−3 In a mixed-wet system, the oil-wet pores correspond to the largest pores in the rock and the small pores are water-wet.2,3 To investigate the effect of wettability, we performed imbibition and drainage displacement tests under water-wet as well as mixed-wet conditions. Another important parameter is interfacial tension (IFT) and its effects on relative permeability. It is known that, at conditions away from the critical point of the system where IFT is high (base IFT), multiphase relative permeability functions may be considered constant, i.e., independent of the flow rate and IFT (immiscible relative permeability functions). At the other extreme (IFT = 0), relative permeability curves reduce to linear functions of the fluid saturation (miscible relative permeability functions). In this work, our focus is “near-miscible” relative permeabilities, which would lie in the region between these two limits. We have studied the behavior of near-miscible gas/oil systems in a wateralternating-gas (WAG) process before.4−8 Recently, we have studied the importance of cyclic hysteresis for gas/oil systems.9 © XXXX American Chemical Society

The aim of the present work is to highlight the impact of some of the pertinent parameters of reservoirs, which should not be ignored during SCAL analysis and reservoir simulations. These are immobile water saturation, saturation history, pore size distribution (permeability), and wettability. In this section, we first define near-miscible displacement and then review the highlights of the literature on the effect of IFT on the shape of relative permeability curves and hysteresis. Also, effects of immobile water saturation and wettability on the performance of gas injection are reviewed. 1.1. Near-Miscible Gas Injection. Under near-miscible conditions, injected gas does not develop complete miscibility with the oil but comes close. Examples of gas injections at these conditions are condensing/vaporizing gas drives, gasfloods at enrichments slightly below minimum miscibility enrichment or at pressures slightly below minimum miscibility pressure.10 Near-critical conditions also occur in volatile-oil or gascondensate systems near their bubble- or dew-point pressures. Many laboratory and some field implementations have indicated that near-miscible and miscible gas injections perform in a comparable manner. However, the general performance can be influenced by a number of physical properties. In many cases, zero IFT is unnecessary, unless the pore size distribution is extremely tight and the rock is oil-wet.11 Oil recovery Received: June 22, 2012 Revised: February 20, 2013

A

dx.doi.org/10.1021/ef301059b | Energy Fuels XXXX, XXX, XXX−XXX

Energy & Fuels

Article

non-wetting (gas) phases. Asar and Handy24 conducted steadystate relative permeability studies for IFTs from 0.82 to 0.01 mN m−1. They found that, as IFT increased, oil relative permeability (wetting phase) decreased more rapidly than the gas (non-wetting phase) relative permeability. Harbert25 found a significant effect of IFT on non-wetting phase relative permeability and a less noticeable effect on wetting phase relative permeability. Henderson et al.26,27 concluded that, as IFT decreases, the oil relative permeability curve remains essentially unaffected, while the relative permeability to gas increases significantly. Schechter and Haynes28 concluded that non-wetting phase relative permeability increases only slightly as the IFT is reduced. The curvature of the wetting phase relative permeability decreased as the IFT was decreasing and straightens at conditions very close to the critical point. Blom et al.29 observed that, at very low IFT, the non-wetting phase relative permeability approaches a unit-slope line for which the non-wetting phase relative permeability is simply equal to its saturation. The wetting phase relative permeability is not affected until the IFT is decreased below 0.06 mN m−1. Al-Wahaibi et al.30 experimental results show that, as the IFT decreases, the non-wetting phase relative permeability increased more rapidly than the wetting phase relative permeability and hysteresis became less important. It should be mentioned that both Al-Wahaibi et al.29 and Blom et al.30 used a liquid/liquid system as a near-miscible system instead of a hydrocarbon liquid/vapor fluid system. The idea of replacement is according to the assumption of the universal behavior of near-critical thermodynamic quantities, for either gas/liquid equilibria or liquid/liquid equilibria in the vicinity of the critical point. Consequently, they assumed that a near-miscible binary liquid system can be used to model a near-miscible gas/liquid system. The main advantage of using a binary liquid system is that experiments can be performed at less extreme conditions than in the case of a gas/liquid system. From the above literature, one can conclude that the knowledge of how one fluid displaces another and how both fluids flow together is needed for physically realistic multiphase numerical simulation. Differences in the wetting preference may account for the difference in flow behavior between different studies. The general result is that flow characteristics for systems with ultralow IFTs ( 5%). For the same positive spreading coefficient, the oil recovery kinetics are different and gravity drainage is much less efficient in the oil-wet sandpack than in the water-wet sandpack. This implies that different displacement mechanisms are involved. In fact, at low oil saturations in water-wet porous medium, the oil forms films on the water covering the solid (spreading oil films), while in oil-wet porous media, oil films are formed directly on the solid (wetting oil films). These two types of films have different characteristics (thicknesses and interactions with the solid) and, thus, different hydraulic conductivities. The results by Dehghan et al.43 of solvent injection (nhexane) showed that the displacement efficiency of the solvents is generally higher in strongly water-wet media. Contrary to these observations, Wylie and Mohanty21 observed that oil recovery is higher in oil-wet media than in water-wet media because of the presence of water-shielding in water-wet media. However, it should be mentioned that the Sw for water-wet and oil-wet conditions was different and mobile in their experiments and recovery mechanisms were based on mass transfer between bypassed oil and gas flow (unequilibrated fluids).

Close to the critical point, any third, non-critical phase will be completely covered by a layer of one of the critical phases. This phenomenon is known as critical-point wetting.31 Consequently, for the water-wet system, in near-miscible gas floods (with zero water saturation), the oil perfectly wets the solid surface (wetting layer). In the presence of immobile water, oil would form a spreading layer on water-wetting layers. Saedi and Handy32 stated that, in the presence of immobile water, the effects of reduced gas/oil IFT may be even more pronounced. Their study showed that the relative permeability ratio of gas/oil shifted to appreciably lower saturations in the presence of immobile water. This indicates more liquid flow at lower liquid saturations (i.e., higher oil relative permeability when immobile water is present). Delclaud et al.23 investigated the effect of immobile water during gas injection (drainage) and stated that water saturation has no effect on gas/oil relative permeabilities. Narahara et al.33 concluded that gas/oil drainage relative permeabilities would not be affected in both water-wet and mixed-wet conditions by the presence of connate water as long as the following conditions are satisfied: (a) connate water is immobile; (b) relative permeabilities are expressed in terms of total liquid saturation; and (c) effective permeability of oil at connate water saturation (instead of absolute permeability for 100% oil saturated) is used as the reference to calculate relative permeabilities. Kalaydjian et al.34 observed that oil recovery and relative permeability are higher for spreading than for non-spreading conditions for gas drainage displacements. Wylie and Mohanty12 observed that, in the presence of water, the MCM gas will tend to have better recovery because many bypassed small pores would contain water as opposed to oil. Skauge et al.35 summarized the effect of connate water on gas/oil displacement. The main effect has been a decrease in oil recovery with decreasing connate water saturation. A further review of the literature36 indicates that, in the presence of connate water, there is generally an increase in oil recovery with water saturation up to a saturation range where water becomes mobile. The higher recovery for the positive spreading system (in comparison to negative spreading) results from two factors:37 (1) flow through the thin but continuous oil films, which is important for mobilizing and reconnecting the waterflood residual oil saturation38 and (2) the oil films avoiding contact with the high curvature regions of the medium and maintaining overall conductivity.39 1.5. Effect of Wettability. The relative distribution of fluids within a porous medium is determined by the wettability of the rock and the IFTs or energies between the three phases. The wettability is very important for the continuity of the phases as wetting films on the solid. These films are in hydraulic continuity over a number of pores even at very low wetting phase saturations. The wetting phase can flow along the solid surface of the pore walls, while the non-wetting phase(s) must flow within the wetting phase. For water-wet systems, the hierarchy of fluid IFTs ensures that water is in direct contact with the rock and any condensed liquid hydrocarbon forms a film on the water surface, separating the gas from the water. However, many oil reservoirs are not strongly water-wet. For a water-wet system, miscibility (development of a zero IFT) does not need to occur to have the optimal gas injection scheme.11 This is due to the fact that, if a system is strongly water-wet and the water adheres more closely to the smaller pore throats, then you do not need or want to enter those pores to recover bulk of the IOIP. For the oil-wet system, where the

2. METHODOLOGY 2.1. Core and Fluid Preparation. The high-pressure, hightemperature core-flooding rig used for the experiments is equipped with an X-ray scanner, which allows us to obtain accurate and repeatable scans of the core before, during, and after the experiments. These scans can be used to obtain the distribution and saturations of different phases (water, oil, and gas) along the core. Core heterogeneities, distribution of immobile water, and front propagation can also be determined and monitored. X-ray results are also used to check for experimental artifacts, such as capillary end effects, which, in our experiments, were not an issue because of the use of long cores and very low IFT (∼0.04 mN m−1) between oil and gas phases. Two different sandstone cores with 1 order of magnitude difference in their absolute permeabilities were used in this study. Table 1 shows the physical properties of these sandstone cores. Prior to performing the coreflood experiments, a core characterization procedure was followed to determine the suitability of the core for kr experiments and also to determine its physical properties. After cleaning and drying, the

Table 1. Physical Properties of Core Samples Used in Experiments

C

core

permeability (mD)

length (cm)

diameter (cm)

porosity (fraction)

C-1 C-2

1000 65

67.1 60.5

4.98 5.08

0.17 0.18

dx.doi.org/10.1021/ef301059b | Energy Fuels XXXX, XXX, XXX−XXX

Energy & Fuels

Article

core was scanned to examine its homogeneity (by porosity profiling) prior to performing the coreflood tests. The core porosity profile was obtained according to the Akin and Kovskec44 formulation. The average porosity of the core obtained from X-ray data is 18.3%, which is comparable to the average value obtained using brine and helium pore volume measurement techniques (18.2%). The brine used in these experiments (as immobile water) was synthetic brine made of sodium chloride (NaCl) and calcium chloride (CaCl2) in degassed and distilled water (Table 2). It should be

the core mixed-wet, whereby parts of the pores remain water-wet (where connate water protects the surfaces of the rock from coming in contact with the crude oil during the aging process) and parts of the pores would be made oil-wet. After the aging period, the core went through another period of cleaning to make sure that the aging crude oil had been displaced from the core and would not contaminate the test fluids. This is a delicate task because the crude oil has to be displaced without aggressive cleaning of the core, which would result in total removal of the adsorbed organic material (which caused wettability alteration) and, hence, rendering ineffective the wettability alteration process. In our experiments, we displaced bulk of crude oil with large pore volume injections of decane. Decane injection continued until the core effluent became clear (no evidence of crude oil). The bulk of decane was removed by high-pressure methane injection, which, in turn, was replaced by the equilibrated oil. Absolute permeability to the equilibrated oil (in the presence of immobile water saturation) has also been measured to make sure that the process of wettability alteration has not affected the connection of the pores and their connectivity. The calculated effective permeability (in the presence of immobile water) for the mixed-wet system was the same as the waterwet system. 2.4. Coreflood Experiments. 2.4.1. Gas Injection (Drainage) Test: 65 mD, Water-Wet, and Swim = 0. One of the displacement types that occur in the reservoir during the WAG process is the displacement of oil by the injected gas slug. To determine gas/oil relative permeability curves applicable to this type of displacement, an unsteady-state gas injection test has been carried out in the 65 mD core. To be able to investigate the impact of immobile water on gas/oil kr and displacement mechanisms, this experiment was carried out with no initial water saturation, Swim. Later, a similar test was carried out with Swim, and the impact of Swim was examined by comparing the results of these tests (Table 4). This experiment started with the core

Table 2. Physical Properties of Synthetic Brine salinity (mg/L)

density at 38 °C (g/L)

viscosity at 38 °C (cP)

1000

992.96

0.68

mentioned that the brine did not represent any particular reservoir brine and was used with the aim of reducing possibilities of adverse reactions between the core and water during primary water imbibition (before immobile water establishment). The hydrocarbon fluid system used in the experiments consisted of an equilibrium binary mixture of methane (C1) and n-butane (nC4). The pressure−temperature (PT) phase diagram of the gas/oil mixture shows a critical point of 1870 psia and 100 °F (38 °C). Because the experiments have been conducted at 1840 psia and 100 °F, the gas and oil were nearly miscible, with the IFT being 0.04 mN m−1 (Table 3). Oil and gas phases were mixed and

Table 3. Measured Fluid Properties for C1−nC4 Binary Mixture at 100 °F and 1840 psia ρg (kg m−3)

ρL (kg m−3)

μg (mPa s)

μL (mPa s)

IFT (mN m−1)

211.4

317.4

0.0249

0.0405

0.04

pre-equilibrated together and also with brine at the conditions of the experiments (1840 psia and 100 °F) to minimize mass transfer between fluids during the displacement experiments. For further details regarding the core and fluid preparation, refer to Fatemi et al.7 2.2. Establishing Immobile Water Saturation (Swim). As mentioned earlier, the process of Swim establishment in the laboratory is a lengthy one because it involves a series of fluid injections and displacements. In this work, the core was dried and cleaned and then water was injected into the core (primary imbibition) to make the core 100% saturated with the brine. Then, mineral oils were used to push the bulk of the brine out of the core and establish the initial immobile water saturation. The mineral oils were displaced using decane (C10), which, in turn, was pushed outside of the core using high-pressure methane (C1). C1 had been previously equilibrated and hydrated with the brine to make sure that there would be no further Sw reduction after mineral oil injection. Finally, C1 was replaced (at test temperature and pressure) with equilibrated oil to initialize the primary stage of the test before the start of the experiments. This procedure resulted in the establishment of Swim = 18%, as obtained by material balance and X-ray scans. X-ray analysis shows that, for both wettability conditions, the established immobile water saturation was homogeneous along the core and the local value was very close to the average value obtained from material balance (18%). Another important feature is that the value and distribution of this immobile water saturation are very close for both wettability conditions. This allows us to directly compare our experimental results from the water-wet sample to those of mixed-wet conditions because they are not affected by saturation history, immobile water saturation value, or its distribution along the core. 2.3. Wettability Alteration Procedure. As mentioned in the previous section, to have a sound comparison between the results of the experiment carried out on the water-wet core to those obtained for the same core after changing its wettability to mixed-wet, it is important to have the same initial water saturation, Swi, in both cores. Hence, we first established the immobile water saturation in the core and then injected crude oil for changing wettability. The core was aged for 2 weeks under 176 °F (80 °C). This process is expected to make

Table 4. Coreflood Experiments Presented in This Work experiment number

core

coreflooding

direction

wettability

Swim

1 2 3 4 5 6 7 8 9

C-1 C-1 C-1 C-2 C-2 C-2 C-2 C-2 C-2

gas injection oil injection gas injection gas injection oil injection gas injection oil injection gas injection oil injection

drainage imbibition drainage drainage imbibition drainage imbibition drainage imbibition

water-wet mixed-wet mixed-wet water-wet water-wet water-wet water-wet mixed-wet mixed-wet

× × ×

× × × ×

fully saturated with the equilibrated oil at 1840 psia and 100 °F. The equilibrated gas was then injected through the core at a constant rate of 100 cm3 h−1, and at the same time, the core effluent was collected in a glass separator (sight glass) at the test pressure (1840 psia). According to the porosity and dimensions of the core, the injection rates applied in this study represent 1.5−3 ft/day. In this range of injection rate, the displacements are quite stable considering very small differences between the viscosities of oil and gas phases. During the experiment, fluid production (oil and gas), injected volumes, and also the pressure at the core inlet and outlet were accurately monitored and measured. For the water-wet system, gas would act as the non-wetting phase with respect to the oil phase. As the gas injection process starts, the gas would displace oil from the larger pores first, because larger pores have smaller capillary pressure to overcome and enter. The conductivity of these pores is also higher compared to the smaller pores, and as a result, the gas would bypass those smaller pores and breakthrough (BT) from the outlet. At the time of BT, the gas would be in the body of the larger pores, while the oil phase is still continues throughout the core as the wetting layers in large pores and untouched in smaller pores. The oil saturation in larger pores would be equal to (or around) D

dx.doi.org/10.1021/ef301059b | Energy Fuels XXXX, XXX, XXX−XXX

Energy & Fuels

Article

oil BT, which happened around 0.5 PV of oil injection, there was a sudden change of slope in the gas recovery curve and the production of gas after BT was negligible (non-wetting phase trapping). 2.4.5. Gas Injection (Drainage) Test: 65 mD, Mixed-Wet, and Swim = 18%. Having altered the wettability of the core, another gas injection was carried out with the same Swim. The results of this mixed-wet experiment were then compared to their water-wet counterpart to examine the effect of wettability. The experiment began with the core containing 82% oil and 18% water (immobile water saturation) at 1840 psia. The rate of oil production from the core slowed after the BT of the gas around 0.35 PV, and thereafter, both the gas and oil were produced. Although the oil recovery rate decreased after the gas BT, it remained significant until the end of the experiment, which is a feature of the near-miscible gas injection process. 2.4.6. Oil Injection (Imbibition) Test: 65 mD, Mixed-Wet, and Swim = 18%. The objective of this experiment was to investigate the impact of wettability on near-miscible gas/oil displacement and relative permeability curves in the imbibition direction. The core was initially saturated with 18% brine and 82% gas at 1840 psia. The oil BT took place around 0.4 PV of oil injection, where there was a sudden change of slope of the gas recovery curve, and afterward, the gas recovery rate dropped to almost zero. 2.4.7. Gas Injection (Drainage) Test: 1000 mD, Water-Wet, and Swim = 8%. The same procedure that was used for establishing the initial water saturation in the 65 mD core was followed for the 1000 mD core too. The value of the established immobile water saturation is less for this core compared to the 65 mD core because of the higher permeability of this core. To be able to compare the results of the water-wet 65 mD core to the results of the 1000 mD core, the test conditions were kept similar (1840 psia and 100 °F, corresponding to a gas/oil IFT value of 0.04 mN m−1). The experiment began by a gas injection in the 1000 mD core containing 92% oil and 8% water (immobile water saturation). The rate of oil recovery from the core slowed after the BT of the gas around 0.46 PV, and thereafter, both the gas and oil were produced. Although the oil recovery rate decreased after the gas BT, it remained significant until the end of the experiment (which is the feature of near-miscible gas injection). 2.4.8. Gas Injection (Drainage) Test: 1000 mD, Mixed-Wet, and Swim = 8%. This gas injection was carried out in the 1000 mD mixedwet core (after changing its wettability) to compare its results to the similar gas injection test that had been carried out in the 65 mD mixed-wet core. A similar wettability alteration procedure as the one used for the 65 mD core sample was also used for the 1000 mD core. A wettability measurement test (USBM) was carried out which showed neutral-wet conditions.47 The test conditions were kept the same for both cores (1840 psia and 100 °F), but in the 1000 mD core, the experiment began with the core initially containing 92% oil and 8% water (immobile water saturation). 2.4.9. Oil Injection (Imbibition) Test: 1000 mD, Mixed-Wet, and Swim = 8%. The objective of this experiment was also to investigate the impact of permeability on near-miscible gas/oil displacement for the imbibition direction. The test conditions were kept similar to the previous oil injection test in the 65 mD mixed-wet rock with initial water saturation. In this test, the core was initially saturated with 8% brine and 92% gas at 1840 psia. The oil BT happened around 0.65 PV of oil injection, where there was a sudden change of slope of the gas recovery curve, at which point the gas recovery rate drops to almost zero.

the residual oil saturation to gas injection, while the oil saturation inside the smaller pores is equal to around 100%. As the process of gas injection continues, considering the very low IFT between oil and gas phases (low capillary entrance), gas would enter smaller pores, displacing the oil phase and making some wetting oil layers. Dependent upon the injection rate, oil/gas IFT, and pore size distribution of the core, BT time and amount of recovery afterward would be different from case to case. The stability of the wetting oil layers, the capillary entrance pressure for smaller pores, and the effectiveness of the gas phase drag forces on the wetting layers would determine whether the gas would produce oil from the oil wetting layers, would prefer to enter new smaller pores, or both. 2.4.2. Oil Injection (Imbibition) Test: 65 mD, Water-Wet, and Swim = 0. The results of the previous experiment in which gas injection was carried out to displace the oil from the core can be used to obtain bounding drainage kr curves for the gas/oil system. For numerical simulation of the WAG process and to consider the hysteresis effect, it is equally important to obtain imbibition gas/oil kr curves as well as drainage.45,46 To determine bounding imbibition gas/oil kr curves for our near-miscible system, another unsteady-state displacement experiment was carried out. The test began by fully saturating the core with the gas (100%) and was carried out at the same pressure and temperature as the previous test (1840 psia and 100 °F) with no initial water saturation. The oil was then injected through the core at a rate of 100 cm3 h−1 (the same injection rate as in the drainage counterpart). Gas production shows that, as expected, prior to oil BT (which occurs around 0.6 PV oil injected), only gas is recovered from the core. However, when the oil front reaches the production end of the core (BT) at 0.6 PV of oil injection, gas production decreases significantly and, thereafter, the core effluent is mostly oil. Continued production of gas (albeit at a very low rate) shows that the oil is unable to completely trap the gas phase after BT and gas production (although at a very small rate) continues along with oil production. In a water-wet rock, oil would act as the wetting phase with respect to the gas phase. As the oil injection process starts, the oil would displace gas from the smaller pores first, because smaller pores have a larger capillary imbibition force for oil. The conductivity of these pores is smaller compared to the larger pores, and as a result, oil relative permeability will be very small. As the oil saturation in these pores increases, wetting phase layers will be thicker, and at high enough saturations, these layers would bridge and trap the gas in between. As the oil injection continues the oil would enter larger pores. The entrapment of the gas inside a pore body is much more difficult in larger pores because oil would have established a continuous path toward the outlet before becoming thick enough to bridge in larger pores. Because the IFT between oil and gas is very low, complete trapping would not occur and a small amount of gas flow and recovery takes place after the BT because of the possible coupled flow of the oil and gas in some medium to larger pores. 2.4.3. Gas Injection (Drainage) Test: 65 mD, Water-Wet, and Swim = 18%. To examine the effect of immobile water saturation (Swim) on the gas/oil relative permeability curves applicable to the displacement of oil by gas (drainage), another unsteady-state gas injection test was carried out. The test conditions were kept similar to the previous gas injection test, with the only difference being the presence of immobile water saturation (Swim) in the core. This experiment began with the core containing 82% oil and 18% water as an immobile phase at 1840 psia. Then, gas injection through the core commenced. During the gas injection period, the rate of injection and production of fluids and the pressure at the inlet and outlet of the core were accurately monitored and measured. Although the oil recovery rate decreased after the gas BT, it remains significant until the end of the experiment (because of near-miscibility conditions). 2.4.4. Oil Injection (Imbibition) Test: 65 mD, Water-Wet, and Swim = 18%. In this experiment, again the impact of immobile water saturation on near-miscible gas/oil relative permeability curves has been investigated, albeit for the imbibition direction. The core initially contained 18% brine and 82% gas at 1840 psia. The gas recovery, oil production, and differential pressure across the core were measured during this oil injection experiment to estimate the kr values. After the

3. RESULTS AND DISCUSSION In this section, we discuss kr determination for the coreflood experiments. Using these experimentally derived kr curves, the effects of hysteresis, immobile water saturation, wettability, and rock permeability (pore size distribution) will be discussed. We will also discuss the differences between gas/oil near-miscible systems with liquid/liquid near-miscible systems. A black-oil core flood simulator (SENDRA) was used in this exercise to history match the core flood results and obtain the best kr estimates. E

dx.doi.org/10.1021/ef301059b | Energy Fuels XXXX, XXX, XXX−XXX

Energy & Fuels

Article

Figure 1. Comparison of (left) displaced phase recovery and (right) pressure drop during two low IFT unsteady-state displacements (imbibition and drainage) (65 mD, water-wet, and Swim = 0).

Figure 2. Simulated oil saturation profile along the core (after history matching) at different stages of the displacement: (left) gas injection and (right) oil injection (65 mD, water-wet, and Swi = 0).

Figure 3. Hysteresis effect on bounding relative permeability drainage and imbibition curves (65 mD, water-wet, and Swim = 0).

3.1. Comparison of Gas and Oil Injection Tests: 65 mD, Water-Wet, and Swim = 0. Figure 1 compares the oil recovery and pressure drop in the near-miscible gas injection (drainage) test to the gas recovery and pressure drop for the near-miscible oil injection (imbibition) test, both carried out in the 65 mD core using the same fluids and performed at the same pressure and temperature. As seen, when oil displaces gas (imbibition), at the BT, the amount of gas recovery is higher than the amount of oil recovery when gas displaces oil (drainage). This means that the oil BT in the oil injection test happened after the gas BT in the gas injection test. However,

extrapolation of the gas and oil recovery curves in Figure 1 reveals that, if the gas injection continues for an extended period of time (in this particular case, more than 3.5 PV), then the amount of oil recovery will catch up with the gas recovery in the oil injection test and would be higher afterward. Figures 1 and 2 show that, although the IFT between oil/gas phases is as low as 0.04 mN m−1, there is still a significant difference between recovery mechanisms, which results in a different recovery profile and pressure drop behavior and, therefore, totally different kr values for imbibition and drainage displacement (see the following sections). The significant additional oil F

dx.doi.org/10.1021/ef301059b | Energy Fuels XXXX, XXX, XXX−XXX

Energy & Fuels

Article

the presence of Swim to that obtained in the absence of Swim (both water-wet). The vertical axis in this figure shows the oil recovery as a fraction of the IOIP, and the horizontal axis is also corrected as the injected volume of the gas in terms of the IOIP pore volume (which assumes the immobile water saturation as a part of the rock). As noted, the presence of Swim has had a positive effect on the gas injection performance because it has delayed the gas BT and has increased the amount of oil recovered during gas injection. A common approach in the simulation of multiphase flow in porous media is that the saturation values are corrected for Swim; i.e., they are reported on the basis of the hydrocarbon pore volume assuming that Swim acts as a part of the rock. The results in Figure 4 clearly show that this assumption is invalid. That is, there is a clear difference in the production profiles in these two plots that have been prepared on the basis of the hydrocarbon pore volume. This observation can be explained in terms of the distribution of different phases at the pore level for these two cases. Because water is the wetting phase, in comparison to the other two phases (oil and gas), it remains in very small pores after initial water establishment. As a result, in both systems, the situation with the larger pores will be the same, with oil in these pores acting as the wetting phase and gas being the non-wetting phase. The differences between two systems stems from the difference between the distributions of different phases in smaller pores. In the absence of immobile water (left panel of Figure 5), the oil would be the wetting phase adhering on the surface of the grains (wetting layer), while the gas, as the nonwetting phase, would flow in the body of the pore. In contrast to this, in the presence of immobile water, water would act as the wetting phase adhering on the grain surfaces (wetting layers) and oil would spread on these wetting layers as the intermediate-wet phase and form spreading layers. Lubricating the flow of oil by water-wetting layers, spreading oil layers (in the presence of immobile water saturation) would be more conductive than oil-wetting layers (in the absence of immobile water saturation). As a result, the performance of gas injection would be improved in the former case, where water-wetting layers exist. This observation is in disagreement with previous reports,23,33,48,49 in which immobile water saturation had no effect on gas/oil displacement and relative permeabilities. This discrepancy can be explained by the very low oil/gas IFT in our experiments. Capillary pressure is a function of IFT, σ, as well as contact angle, θ, and pore radius, r, through the following equation:

recovery after the gas BT in near-miscible gas injection is attributed to the flow and recovery of the oil through highly conductive oil layers that exist in near-miscible systems. In nearmiscible oil injection experiments, however, because the gas is a non-wetting phase, no gas layers can form, and hence, the gas left behind after the BT of the oil would be mainly isolated and fragmented. In the vicinity of miscibility, these two recovery curves and pressure drop behavior would approach together, in the case of two liquid phases.29,30 This means that, whether the wetting phase displaces the non-wetting phase or vice versa, recovery curves and pressure drop behavior would be very close for the liquid/liquid near-miscible system. This is due to the fact that, in the case of the liquid/liquid system, the amount of trapped non-wetting phase in imbibition would be very small and close to the residual wetting phase in drainage displacement. Considering this, Figure 1 highlights the difference between the oil/gas system and the liquid/liquid29,30 system at near-miscible conditions. As a result, the application of the near-miscible liquid/liquid system instead of the gas/liquid system to investigate near-miscible gas injections, gas/ condensate, or gas/volatile-oil systems is not justified and would result in erroneous conclusions and results. The liquid/ liquid system has been usually applied in the literature because of the simplicity of the experiments in liquid/liquid systems compared to gas/liquid (oil) systems (which would require working under high pressures). We believe for the same low IFT value (near-miscible conditions) that the behavior of a liquid/liquid system and the associated displacement mechanisms is different from those of a gas/liquid system. This significant difference in the behavior of the two systems is responsible for the significant kr hysteresis that we have observed (Figure 3) for our near-miscible gas/oil system compared to the experiments by Al-Wahaibi et al.29 and Blom et al.,30 which were performed on a liquid/liquid near-miscible system. We have also observed the same behavior (significant trapping of the non-wet phase during imbibition) in our nearmiscible gas/oil displacement experiments in the mixed-wet (65 mD) core when the core contained immobile water. Our experimental results in the higher permeability core (1000 mD) have also confirmed these observations. 3.2. Effect of Swim on Gas Injection (Drainage): 65 mD and Water-Wet. Figure 4 shows the effect of immobile water saturation (Swim) on the gas injection performance by comparing the oil recovery curve obtained by gas injection in

Pcog =

2σog cos θ r

For high oil/gas IFT systems,23,33,49,50 capillary forces (between oil and gas) would be high enough to prevent the gas entering into small pores. As a result, in a high oil/gas IFT system, in both cases with and without immobile water saturation, gas would not be able to enter those very small pores anyway and the amount of oil recovered by gas and kr would be unaffected by the presence or absence of immobile water saturation. In our system (very low IFT), however, because of very low oil/gas IFT, oil/gas capillary pressures would be low and the gas phase would be able to enter small pores (in the absence or presence of immobile water). Taking into account the difference between the pore level distributions of the fluids in the absence or

Figure 4. Effect of Swim on oil recovery (fraction of IOIP) as a function of injected gas (fraction of the IOIP pore volume) (65 mD, water-wet, and drainage). G

dx.doi.org/10.1021/ef301059b | Energy Fuels XXXX, XXX, XXX−XXX

Energy & Fuels

Article

Figure 5. Schematics of pore level distribution of different phases (oil, water, and gas) in small pores of a rock (left) in the absence of immobile water saturation and (right) in the presence of immobile water saturation (brown, grains; black, oil; blue, water; and red, gas) (65 mD, water-wet, and drainage).

Figure 6. Effect of Swim on oil and gas relative permeabilities (left) versus total liquid saturation and (right) versus oil saturation (65 mD, water-wet, and drainage).

presence of Swim, the amount of oil recovery and kr values would be affected. The extent of the oil recovery improvement in the presence of immobile water saturation would be a function of the pore size distribution and oil/gas IFT. Figure 6 shows the effect of immobile water on oil and gas relative permeability values for drainage displacement. Although in the left panel of Figure 6, the conditions proposed by Narahara et al.33 are satisfied (a, connate water is immobile; b, relative permeabilities are expressed in terms of total liquid saturation; and c, effective permeability of oil at immobile water saturation is used as a reference to calculate relative permeabilities), relative permeability is different with and without immobile water saturation. In the left panel of Figure 6 (kr expressed in terms of liquid saturation), oil and gas relative permeability values are less in the presence of immobile water, because for the same liquid saturation, part of the pore space is now filled with the water phase, which is immobile, and as a result, the relative conductance of the core has been reduced. In the right panel of Figure 6, where oil and gas relative permeabilities are expressed in terms of the oil saturation, kro shows improvement for the case when immobile water is present, while krg values have been reduced in this case. This can be explained by the fact that, for the same oil saturation value, the conductance of the oil layers are more in spreading conditions (in the presence of immobile water saturation) compared to the case of wetting layers (absence of Swim). 3.3. Effect of Swim on Oil Injection (Imbibition): 65 mD and Water-Wet. Figure 7 depicts the effect of Swim on the oil

Figure 7. Gas production (fraction of IGIP) as a function of injected oil (fraction of the IGIP pore volume) (65 mD, water-wet, and imbibition).

injection performance by comparing the gas recovery curve obtained by oil injection in the presence of Swim to that obtained in the absence of Swim. The vertical axis in this figure shows the gas recovery as a fraction of the initial gas in place (IGIP), and the horizontal axis is also corrected to represent the injected volume of the oil in terms of the IGIP pore volume. As noted, the presence of Swim has had a negative effect on the oil injection performance. The results in Figure 7 clearly show that the hypothesis of assuming immobile water as a part of rock is invalid. This observation can be explained in terms of the pore-scale distribution of fluid phases for these two H

dx.doi.org/10.1021/ef301059b | Energy Fuels XXXX, XXX, XXX−XXX

Energy & Fuels

Article

Figure 8. Schematics of the pore level distribution of different phases in small pores of the core sample (left) in the absence of immobile water saturation and (right) in the presence of immobile water saturation (brown, grains; black, oil; blue, water; and red, gas) (65 mD, water-wet, and imbibition).

Figure 9. Effect of Swim on oil and gas relative permeabilities (left) versus total liquid saturation and (right) versus oil saturation (65 mD, water-wet, and imbibition).

Figure 10. Hysteresis effect on bounding relative permeability drainage and imbibition curves (65 mD, water-wet, and Swim = 18%).

experiments. As shown in Figure 8, for the entrapment of the gas phase in the body of the pore (because of bridging of the oil layers at the throats and the snap-off process), lower oil saturation (less oil) would be required when part of the pore is occupied by water (immobile water saturation). As a result, the oil BT happens earlier compared to the case in which immobile water is absent. Figure 9 shows the effect of immobile water saturation on the oil and gas relative permeability values for imbibition displacement. Although in the left panel of Figure 9, the conditions proposed by Narahara et al.33 are satisfied, the relative permeability is still different as a result of the absence or presence of immobile water. Similar to what has already been

explained for the case of drainage, in the left panel of Figure 9 (kr expressed in terms of liquid saturation), oil and gas relative permeabilities are less in the presence of immobile water, because for the same liquid saturation, part of the pore space is now filled with the water phase, which is immobile, and as a result, the flow capacity of the pore space has been reduced. In the right panel of Figure 9, where oil and gas relative permeabilities are expressed in terms of oil saturation, kro shows improvement in the presence of immobile water, while the krg values have been reduced in this case. The data shown in Figures 6 and 9 also show that the effect of immobile water saturation is more pronounced for the non-wetting phase (gas) compared to the wetting phase (oil). Figure 10 shows the oil I

dx.doi.org/10.1021/ef301059b | Energy Fuels XXXX, XXX, XXX−XXX

Energy & Fuels

Article

spreading layers (left panel of Figure 12). For the mixed-wet system, parts of these pores (which have been in contact with crude oil during the wettability alteration process) are neutral to oil-wet (right panel of Figure 12). In these pores, oil layers would change their action from spreading layers to the wetting layers and vice versa. As discussed earlier, the conductivity of wetting layers are less than that of spreading layers, and as a result, oil recovery in the mixed-wet system would be less compared to that in the water-wet case. Figure 13 shows the effect of wettability on the oil and gas relative permeability curves for the experiments with immobile water saturation. As seen from this figure, oil relative permeability has decreased for the mixed-wet core compared to the water-wet system. This can be explained using the same discussion that has already been provided for oil recovery. The same figure shows that the gas relative permeability has also decreased for the mixed-wet system compared to the water-wet sample. Figure 14 shows the pore level distribution of different phases in the small pores of the mixed-wet system. As mentioned, the conductivity of the wetting oil layers would be less than the spreading oil layers. The oil saturation increases at these points (non-water-wet pores), and there is a chance of the layers bridging if the oil saturation is high enough (or pore throats are small enough). These bridges are not stable because local oil saturation at this pore would decrease by feeding the adjacent spreading oil layers and, as a result, would collapse again (because of very low IFT and capillary pressure between oil and gas phases). Although the gas would not be trapped inside the pore body as a result of this evidence, its relative permeability would decrease. For further discussions regarding the formation and collapse of these liquid bridges and their possible effects on relative permeability, refer to refs 19 and 50. 3.5. Effect of Wettability on Oil Injection. A similar comparison to what was carried out above for the gas injection (drainage) experiments has been carried out in this section to investigate the effect of wettability alteration on the performance of oil injection (imbibition). Figure 15 compares gas recovery by oil injection in the 65 mD core for the water-wet and mixed-wet systems. The oil BT in the water-wet rock is slightly delayed in comparison to that in the mixed-wet rock. The same graph also shows that the ultimate oil recovery factor achieved for the mixed-wet core is lower compared to the case of the water-wet core. In the mixed-wet core, the presence of partly wetting oil layers and partly spreading layers reduces the conductivity of the oil layers compared to the water-wet system, in which the oil phase is connected all of the way through the

and gas relative permeability curves for imbibition and drainage for the tests with immobile water saturation. When this figure is compared to Figure 3, it can be concluded that the hysteresis effect is less when immobile water is present in the rock. In the pores in which water exists, oil would act as an intermediatewet phase (in comparison to its wetting behavior in the absence of immobile water), and as a result, the hysteresis effect would be less. 3.4. Effect of Wettability on the Gas Injection Process. To investigate the possible effects of the rock wettability on the performance of gas injection, the experimental results of oil recovery by gas injection in the water-wet and mixed-wet cores (for the same Swim) have been compared. Figure 11 shows that

Figure 11. Effect of wettability on the performance of gas injection (65 mD and Swi = 18%).

the gas BT in the water-wet rock takes place slightly later than that in the mixed-wet core. The same graph also shows that the oil production rate and the ultimate recovery factor achievable in the mixed-wet core are lower compared to those in the water-wet sample. Both of these observations indicate that the performance of gas injection has been adversely affected in a less water-wet system. In the mixed-wet core, the pores can be divided in two categories, partly water-wet and partly intermediate- to oil-wet. The performance of the very large pores (in which there is no immobile water) would be more or less the same for both systems, because oil would be the wetting phase anyway (oil recoveries at the BT are very similar for both systems). The difference would be more pronounced for smaller pores, in which immobile water is present. For water-wet systems, oil would be connected through the

Figure 12. Pore level distribution of different phases in small pores of the core sample: (left) water-wet system and (right) mixed-wet system (brown, grains; black, oil; blue, water; and red, gas) (65 mD, drainage, and Swim = 18%). J

dx.doi.org/10.1021/ef301059b | Energy Fuels XXXX, XXX, XXX−XXX

Energy & Fuels

Article

Figure 13. Effect of wettability on oil and gas relative permeability curves (65 mD, drainage, and Swim = 18%).

Figure 14. Pore level distribution of different phases in small pores of the mixed-wet core sample. From left to right shows the temporary formation of oil bridges, which would collapse again as the gas injection and oil production continues (brown, grains; black, oil; blue, water; and red, gas) (65 mD, mixed-wet, and drainage).

highly conductive spreading layers. A lower conductivity of the layers in mixed-wet systems results in the accumulation of the oil in locations where the oil phase is present as wetting layers. As the oil injection continues and the oil saturation inside the core increases, there is a high chance for these layers to bridge at the pore throats and snap-off the gas phase inside the pore body (Figure 16). For the case of water-wet rock, because of the high conductivity of the oil spreading layers, a larger oil saturation (in comparison to the mixed-wet system) would be required to thicken these layers and bridge at the pore throats (Figure 17). As a result, the process of gas snap-off would be delayed for water-wet systems compared to mixed-wet systems, and higher gas recovery would be achieved. Figure 18 shows the effect of wettability on oil and gas relative permeabilities in the presence of immobile water.

Figure 15. Effect of wettability on the performance of the oil injection test (65 mD, imbibition, and Swim = 18%).

Figure 16. Pore level distribution of different phases in small pores of the mixed-wet core sample. From left to right shows the oil saturation increment and formation of stabilized oil bridges, which would snap-off the gas phase inside the pore body (brown, grains; black, oil; blue, water; and red, gas) (65 mD, mixed-wet, and imbibition). K

dx.doi.org/10.1021/ef301059b | Energy Fuels XXXX, XXX, XXX−XXX

Energy & Fuels

Article

Figure 17. Pore level distribution of different phases in small pores of the water-wet core sample. From left to right shows the oil saturation increment. Because the spreading layers are very conductive, formation of stabilized oil bridges (and, as a result, gas snap-off) would be delayed to larger oil saturations compared to the mixed-wet system (brown, grains; black, oil; blue, water; and red, gas) (65 mD, water-wet, and imbibition).

Figure 18. Effect of wettability on oil and gas relative permeability curves (65 mD, imbibition, and Swim = 18%).

3.6. Effect of Rock Permeability. Figure 19 shows the effect of wettability on the performance of gas injection for the

4. CONCLUSION (1) Current assumptions in the literature suggest that, as a fluid system approaches miscibility, relative permeabilities of both phases, as a function of the wetting phase saturation, show linear behavior and become diagonal lines. The results of this study show that, only for the non-wetting phase (gas), in the drainage direction, this assumption is valid. However, for the two rocks tested in this work, the wetting phase relative permeability (oil), in both imbibition and drainage directions, and also the non-wetting phase (gas) kr, in the imbibition direction, show significant deviation from this assumption even at the very low IFT of 0.04 mN m−1. (2) On the basis of the current assumptions in the literature, as we approach miscibility, non-wetting phase relative permeability hysteresis diminishes and, normally, no hysteresis is assumed for the wetting phase. The results presented here show that, even at near-miscible conditions (IFT = 0.04 mN m−1), we have significant hysteresis for both the non-wetting phase (gas) and the wetting phase (oil), in both the high permeability (1000 mD) and low permeability (65 mD) rocks. (3) On the basis of our results, low IFT liquid/liquid systems should not be used as an analogue for low IFT gas/liquid (oil) systems. The behavior of these two systems and the involved displacement mechanisms are different. (4) In agreement with literature data, hysteresis was much higher for the non-wetting phase (gas) compared to that for the wetting phase (oil). (5) Hysteresis for both the wetting and non-wetting phases was less in the highly permeable core (1000 mD) compared to that in the 65 mD core sample. (6) Current assumptions in the literature suggest that, as long as connate water is immobile, it does not influence the relative permeability of the phases. Our

Figure 19. Effect of wettability on the performance of gas injection (1000 mD, drainage, and Swim = 8%).

1000 mD core. Comparing this figure to Figure 11 (the 65 mD core) shows that the effect of wettability is not significant in 1000 mD. This is attributed to the higher permeability and larger pore sizes in the 1000 mD core, which decrease the effect of capillary forces. Figure 20 shows the hysteresis effect on oil and gas relative permeability bounding curves for the 1000 mD mixed-wet rock. When this figure is compared to Figure 10, which shows the corresponding kr for the 65 mD, mixed-wet core, one can conclude that, although the hysteresis effect is less for the high permeability (1000 mD) core, it cannot be ignored. L

dx.doi.org/10.1021/ef301059b | Energy Fuels XXXX, XXX, XXX−XXX

Energy & Fuels

Article

Figure 20. Effect of hysteresis on bounding relative permeability curves in the mixed-wet system for the 1000 mD rock (drainage, blue; imbibition, red). the SPE Annual Technical Conference and Exhibition; Dallas, TX, Oct 1−4, 2000; SPE-63000. (5) Sohrabi, M.; Tehrani, D. H.; Danesh, A.; Henderson, G. D. Visualisation of oil recovery by water alternating gas (WAG) injection using high pressure micromodelsOil-wet and mixed-wet systems. Proceedings of the SPE Annual Technical Conference and Exhibition; New Orleans, LA, Sept 30−Oct 3, 2001; SPE-71494. (6) Sohrabi, M.; Tehrani, D. H.; Danesh, A.; Henderson, G. D. Visualization of oil recovery by water-alternating-gas injection using high-pressure micromodels, SPE-89000. SPE J. 2004, 9 (3), 290−301. (7) Fatemi, S. M.; Sohrabi, M. Experimental investigation of nearmiscible water-alternating-gas injection performance in water-wet and mixed-wet systems. SPE J. 2013, 18 (1), 114−123. (8) Sohrabi, M.; Fatemi, S. M. Experimental investigation of oil recovery by different injection scenarios under low oil/gas IFT and mixed-wet condition: Water-flood, gas injection, WAG and SWAG injection. Proceedings of the Abu Dhabi International Petroleum Exhibition and Conference; Abu Dhabi, United Arab Emirates (UAE), Nov 11−14, 2012; SPE-161074. (9) Fatemi, S. M.; Sohrabi, M.; Jamiolahmady, M.; Ireland, S. Experimental and theoretical investigation of gas/oil relative permeability hysteresis under low oil/gas interfacial tension and mixed-wet conditions. Energy Fuels 2012, 26 (7), 4366−4382. (10) Shyeh-Yung, J. J.; Stadler, M. P. Effect of injection composition and pressure displacement of oil by enriched hydrocarbon gases. SPE Reservoir Eng. 1995, 10 (2), 109−115. (11) Thomas, F. B.; Erian, A.; Zhou, X.; Bennion, D. B.; Bennion, D. W.; Okazawa, T. Does miscibility matter in gas injection? Proceedings of the Petroleum Society of Canada Annual Technical Meeting; Calgary, Alberta, Canada, June 7−9, 1995. (12) Wylie, P.; Mohanty, K. K. Effect of water saturation on oil recovery by near-miscible gas injection. SPE Reservoir Eng. 1997, 12 (4), 264−268. (13) Burger, J. E.; Bhogeswara, R.; Mohanty, K. K. Effect of phase behavior on bypassing in enriched gasfloods. SPE Reservoir Eng. 1994, 9 (2), 112−118. (14) Thomas, F. B.; Holowach, N.; Zhou, X.; Bennion, D. B.; Benion, D. W. Miscible or near-miscible gas injection, which is better?. Proceedings of the SPE/DOE Symposium on Improved Oil Recovery; Tulsa, OK, April 17−20, 1994; SPE 27811. (15) Pande, J. B.; Orr, F. M., Jr. Analytical computation of breakthrough recovery for CO2 floods in layered reservoirs. Proceedings of the SPE/DOE Symposium on Enhanced Oil Recovery; Tulsa, OK, April 22−25, 1990; SPE 20177. (16) Sohrabi, M.; Danesh, A.; Tehrani, D. H.; Jamiolahmady, M. Microscopic mechanisms of oil recovery by near-miscible gas injection. Transp. Porous Media 2008, 72 (3), 351−367. (17) Sohrabi, M.; Danesh, A.; Jamiolahmady, M. Visualisation of residual oil recovery by near-miscible gas and SWAG injection using

investigation shows that the effect of connate water on wetting phase relative permeability (oil) is not significant if the kr data are presented on the basis of the oil (wetting phase) saturation. However, the effect of connate water saturation on the nonwetting phase (gas) was significant and should not be ignored. This shows the importance of performing SCAL (kr) tests with representative connate water saturation. More studies regarding the effect of connate water saturation are recommended. (7) Relative permeability of both the wetting and non-wetting phases (oil and gas) in the mixed-wet systems reduced compared to that in the water-wet systems. The reduction was observed for both imbibition and drainage directions. This shows the importance of performing SCAL tests under representative wettability of the reservoir rock (for nearmiscible conditions and in the presence of immobile water saturation).

■

AUTHOR INFORMATION

Corresponding Author

*Telephone: +44-(0)131-451-3198. E-mail: mobeen.fatemi@ pet.hw.ac.uk. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was carried as part of the ongoing Three-Phase Flow and Water-Alternating-Gas (WAG) Injection Joint Industry Project (JIP) at Heriot-Watt University. The project is supported equally by the UK DECC, BHP Billiton, Statoil ASA, Petrobras, Dong Energy, BG Group, Total E&P, and Chevron Upstream Europe, which is gratefully acknowledged. The authors also acknowledge Shaun Ireland for his help during the coreflood experiments.

■

REFERENCES

(1) Jerauld, G. R.; Rathmell, J. J. Wettability and relative permeability of Prudhoe Bay: A case study in mixed-wet reservoirs. SPE Reservoir Eng. 1997, 12 (1), 58−65. (2) Salathiel, R. A. Oil recovery by surface film drainage in mixedwettability rocks. J. Pet. Technol. 1973, 25 (10), 1216−1224. (3) Delshad, M.; Lenhard, R. J.; Oostrom, M.; Pope, G. A. A mixedwet hysteretic relative permeability and capillary pressure model for reservoir simulations. SPE Reservoir Eval. Eng. 2003, 6 (5), 328−334. (4) Sohrabi, M.; Henderson, G. D.; Tehrani, D. H.; Danesh, A. Visualisation of oil recovery by water alternating gas (WAG) injection using high pressure micromodelsWater-wet system. Proceedings of M

dx.doi.org/10.1021/ef301059b | Energy Fuels XXXX, XXX, XXX−XXX

Energy & Fuels

Article

high-pressure micromodels. Transp. Porous Media 2008, 74 (2), 239− 257. (18) Hartman, K. J.; Cullick, A. S. Oil recovery by gas displacement at low interfacial tension. J. Pet. Sci. Eng. 1994, 10, 197−210. (19) Williams, J. K.; Dawe, R. A. Photographic observation of unusual flow phenomena in porous media at interfacial tensions below 0.1 mN/m. J. Colloid Interface Sci. 1988, 124 (2), 691−696. (20) Jamiolahmady, M.; Danesh, A.; Tehrani, D. H.; Duncane, D. B. A mechanistic model of gas-condensate flow in pores. Transp. Porous Media 2000, 41, 17−46. (21) Hubbert, M. K. Darcy’s law and the field equations of the flow of underground fluids, SPE-749-G. Pet. Trans. AIME 1956, 207, 222− 239. (22) Bardon, C.; Longeron, D. G. Influence of very low interfacial tensions on relative permeability. SPE J. 1980, 20 (5), 391−401. (23) Delclaud, J.; Rochon, J.; Nectoux, A. Investigation of gas/oil relative permeabilities: High permeability oil reservoir application. Proceedings of the SPE Annual Technical Conference and Exhibition; Dallas, TX, 1987; SPE 16966. (24) Asar, H.; Handy, L. L. Influence of interfacial tension on gas/oil relative permeability in a gas-condensate system. SPE Reservoir Eng. 1988, 3 (1), 257−264. (25) Harbert, L. W. Low interfacial tension relative permeability. Proceedings of the SPE Annual Technical Conference and Exhibition; San Francisco, CA, Oct 5−8, 1983; SPE-12171. (26) Henderson, G. D.; Danesh, A.; Tehrani, D. H.; Peden, J. M. The effect of velocity and interfacial tension on relative permeability of gas condensate fluids in the wellbore region. J. Pet. Sci. Eng. 1997, 17 (3− 4), 265−273. (27) Henderson, G. D.; Danesh, A.; Tehrani, D. H.; Al-Shaidi, S.; Peden, J. M. Measurement and correlation of gas condensate relative permeability by the steady-state method, SPE-30770. SPE Reservoir Eval. Eng. 1998, 1 (2), 134−140. (28) Schechter, D. S.; Haynes, J. M. Relative permeabilities of a near critical binary fluid. Transp. Porous Media 1992, 9 (3), 241−260. (29) Blom, S. M. P.; Hagoort, J.; Soetekouw, D. P. N. Relative permeability at near-critical conditions. SPE J. 2000, 5 (2), 172−181. (30) Al-Wahaibi, Y. M.; Grattoni, C. A.; Muggeridge, A. H. Drainage and imbibition relative permeabilities at near miscible conditions. J. Pet. Sci. Eng. 2006, 53 (3−4), 239−253. (31) Cahn, J. W. Critical point wetting. J. Chem. Phys. 1977, 66 (8), 3667−3672. (32) Saeidi, A.; Handy, L. L. Flow and phase behavior of gas condensate and volatile oils in porous media. Proceedings of the SPE California Regional Meeting; San Francisco, CA, April 4−5, 1974; SPE 4891. (33) Narahara, G. M.; Pozzi, A. L.; Blackshear, T. H. Effect of connate water on gas/oil relative permeabilities for water-wet and mixed-wet Berea rock. SPE Adv. Technol. Ser. 1993, 1 (2), 114−122. (34) Kalaydjian, F.; Vizika, O.; Moulu, J.; Munkerud, P. K. The role of wettability and spreading in gas injection processes under secondary conditions. In New Developments in Improved Oil Recovery; De Hann, H. J., Ed.; Geological Society of London: London, U.K., 1995; Geophysical Society Special Publication Number 84, pp 63−71. (35) Skauge, A.; Eleri, O. O.; Graue, A.; Monstad, P. A. Influence of connate water on oil recovery by gravity drainage. Proceedings of the SPE/DOE 9th Symposium on Enhanced Oil Recovery; Tulsa, OK, April 17−20, 1994; SPE/DOE 27817. (36) Skauge, A.; Ottesen, B. A summary of experimentally derived relative permeability and residual saturation on North Sea reservoir cores. Proceedings of the International Symposium of the Society of Core Analysts; Monterey, CA, Sept 23−35, 2002; SCA2002-12. (37) Agbalaka, C.; Dandekar, A. Y.; Patil, S. L.; Khataniar, S.; Hemsath, J. R. The effect of wettability on oil recovery: A review. Proceedings of the SPE Asia Pacific Oil and Gas Conference and Exhibition; Perth, Western Australia, Australia, Oct 20−22, 2008; SPE114496.

(38) Oren, P. E.; Billiotte, J.; Pinczewski, W. V. Mobilization of waterflood residual oil by gas injection for water-wet conditions. SPE Form. Eval. 1992, 7 (1), 70−78. (39) Vizika, O.; Lombard, J. M. Wettability and spreading: Two key parameters in oil recovery with three-phase gravity drainage. SPE Reservoir Eng. 1996, 11 (1), 54−60. (40) Chatzis, I.; Kantzas, A.; Dullien, F. A. L. On the investigation of gravity-assisted inert gas injection using micromodels, long Berea sandstone cores, and computer-assisted tomography. Proceedings of the SPE Annual Technical Conference and Exhibition; Houston, TX, Oct 2− 5, 1998; SPE 18284. (41) Kalaydjian, F. J.-M. Performance and analysis of three-phase capillary pressure curves for drainage and imbibition in porous media. Proceedings of the SPE Annual Technical Conference and Exhibition; Washington, D.C., Oct 4−7, 1992; SPE 24878. (42) Oren, P. E.; Pinczewski, W. V. The effect of wettability and spreading coefficients on the recovery of waterflood residual oil by miscible gasflooding, SPE 24881. SPE Form. Eval. 1994, 9 (2), 149− 156. (43) Dehghan, A. A.; Kharrat, R.; Ghazanfari, M. H.; Farzaneh, S. A. Studying the effects of pore geometry, wettability and co-solvent types on the efficiency of solvent flooding to heavy oil in five-spot models. Proceedings of the Asia Pacific Oil and Gas Conference and Exhibition; Jakarta, Indonesia, Aug 4−6, 2009; SPE-123315. (44) Akin, S.; Kovscek, R. Computed tomography in petroleum engineering research. In Applications of X-ray Computed Tomography in the Geosciences; Geological Society of London: London, U.K., 2003; Geological Society Special Publication Number 215, pp 23−38. (45) Spiteri, E. J.; Juaned, R. Impact of relative permeability hysteresis on the numerical simulation of WAG injection. J. Pet. Sci. Eng. 2006, 50 (2), 115−139. (46) Hustad, O. S.; Browning, D. J. A fully coupled three-phase model for capillary pressure and relative permeability for implicit compositional reservoir simulation, SPE-125429. SPE J. 2010, 15 (4), 1003−1019. (47) Sohrabi, M.; Tehrani, D. H.; Al-Abri, M. Performance of nearmiscible gas and SWAG injection in a mixed-wet core. Proceedings of the International Symposium of the Society of Core Analysts; Calgary, Alberta, Canada, Sept 10−12, 2007; SCA 2007-26. (48) Owens, W. W.; Parrish, D. R.; Lamoreaux, W. E. An evaluation of gas drive method for determining relative permeability relationship. Pet. Trans. AIME 1956, 207, 275−280. (49) Knopp, C. R. Gas−oil relative permeability ratio correlation from laboratory data. J. Pet. Technol. 1956, 17 (9), 1111−1122. (50) Dawe, R. A.; Grattoni, C. A. Fluid flow behaviour of gascondensate and near-miscible fluids at the pore scale. J. Pet. Sci. Eng. 2006, 55 (3−4), 228−236.

N

dx.doi.org/10.1021/ef301059b | Energy Fuels XXXX, XXX, XXX−XXX