NMR Studies of the Effect of CO2 on Oilfield Emulsion Stability

Jun 6, 2016 - Ionic strength-dependent pre-asymptoic diffusion coefficient distribution in porous media - Determination through the pulsed field gradi...
0 downloads 10 Views 4MB Size
Download PDF
Article pubs.acs.org/EF

NMR Studies of the Effect of CO2 on Oilfield Emulsion Stability Nicholas N. A. Ling, Agnes Haber, Thomas J. Hughes, Brendan F. Graham, Eric F. May, Einar O. Fridjonsson, and Michael L. Johns* Fluid Science and Resources Division, School of Mechanical and Chemical Engineering, The University of Western Australia, 35 Stirling Highway, Crawley, Western Australia 6009, Australia ABSTRACT: Formation of water-in-crude oil emulsions is a pervasive problem for crude oil production and transportation. Here we investigate the effectiveness of a comparatively low pressure CO2 treatment in terms of breaking these water-in-crude oil emulsions. To this end, we used unique benchtop nuclear magnetic resonance (NMR) technology to measure the droplet size distribution (DSD) of the emulsions. Treatment with 50 bar CO2 for 2 h resulted in significant emulsion destabilization; this was replicated when CO2 was replaced by N2O, which has a solubility in both the aqueous and oil phases similar to that of CO2. Low solubility gases, N2 and CH4, by contrast had no effect on emulsion stability. Treatment with CO2 was also found to have no effect on a model water-in-paraffin oil emulsion stabilized by a synthetic surfactant (Span 80). Collectively, this supported the hypothesis that emulsion destabilization results from CO2 precipitation of asphaltenes as opposed to emulsion droplet film disruption during depressurization, which are the two competing theories reported in the literature to explain the observed supercritical CO2 destabilization of emulsions. Treatment of a water-in-crude oil emulsion featuring partial removal of asphaltenes from the oil phase was consistent with this hypothesis, as the effect of the CO2 treatment on emulsion destabilization was significantly more pronounced.

1. INTRODUCTION Through the lifetime of an oilfield, the water cut can increase to in excess of 95 vol %.1 Crude oil, along with formation water (brine solution) from the reservoir, must be transported through wellheads, pumps, pipelines, and other processing facilities. The vigorous agitation and shearing experienced throughout this transport often causes emulsions to be formed, mainly water-in-crude-oil emulsions.2 Such emulsions can cause many problems; in particular, they are difficult to transport and can be very difficult to separate (break) into distinct water and crude oil phases during the required downstream processing.3,4 Such separation of the emulsified water is, however, usually critical to meet the sales specifications of the crude oil.5 These water-in-crude oil emulsions are commonly stabilized by naturally occurring surface-active agents (surfactants) and fine particles. Surfactants are molecules with hydrophilic and hydrophobic moieties, allowing them to concentrate at the oil/ water interface, stabilizing the emulsion droplets and reducing the energy required to form emulsions by reducing the interfacial tension.6 Natural surfactants found in crude oil are varied and consist of asphaltic materials,7 resinous substances, and oil-soluble organic acids.8 Emulsions can also be mechanically stabilized by fine particles: these are much smaller than the emulsion droplets and can be clay particles, waxes, mineral scales, corroded materials, or drilling muds.5 Factors such as particle size, interaction between particles, and wettability of the particles allow certain fine solids to collect at the oil/water interface and stabilize the emulsions.9 Several methods of breaking water-in-crude oil emulsions are employed in industry including the addition of demulsifier chemicals, heating, distillation, filtering, and gravitational settling.10 However, these methods are generally expensive, ineffective, and often not environmental friendly. Demulsifier chemicals work by counteracting the stabilizing effect of © XXXX American Chemical Society

surfactants, often by replacing them at the water−oil interface. Typical demulsifiers are the polymeric chains of ethylene oxides, poly(propylene oxide)s of alcohol, ethoxylated phenols, ethoxylated alcohols and amines, ethoxylated resins, ethoxylated nonylphenols, polyhydric alcohols, and sulfonic acid salts.5 However, crude oil chemistry can change significantly through the lifetime of a reservoir meaning that the composition of demulsifier treatments often requires frequent revision.11 Heating is applied to lower the emulsion viscosity and is often used in tandem with demulsifiers. It can also weaken the interfacial film due to thermal expansion of the droplets, hence enabling greater droplet coalescence. Heat application of course requires significant energy input and can also result in the undesirable loss of lighter hydrocarbon content. High energy consumption, variable effectiveness, and complexity also often characterize emulsion breaking methods based on distillation, filtering, and gravitational settling.12 In the work presented here, we consider the use of comparatively low pressure CO2 gas (20−50 bar) to break water-in-crude-oil emulsions. Such a treatment is cheap, as it can readily employ existing waste gas streams (which are possibly already at an elevated pressure); it does not contaminate any (depressurized) discharge water or oil products and avoids the use of supercritical fluids and hence associated problems with corrosion. In terms of the effect of CO2 on emulsion stability, work conducted thus far, and reported in the open literature, has exclusively focused on the application of supercritical CO2. Generally, the results have been interpreted in terms of two main mechanistic theories. Received: April 4, 2016 Revised: May 30, 2016

A

DOI: 10.1021/acs.energyfuels.6b00782 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels Theory 1, patented by Sjoblom et al.,13 involves dissolution of the CO2 into the aqueous (droplet) phase at elevated (supercritical) pressures, followed by depressurization and formation of CO2 bubbles that rupture the water−oil interface stabilized by surfactants.14−16 Consequently, droplet coalescence is facilitated, which ultimately leads to the breaking of the emulsion. Theory 2, proposed by Zaki et al.,17 states that asphaltenes precipitate during CO2 saturation at elevated pressure, causing removal of asphaltenes from the water−oil interface and hence resulting in the emulsion becoming unstable. This asphaltene precipitation phenomenon is widely reported and discussed in the literature.18−20 Nuclear magnetic resonance (NMR) measurements have been demonstrated to be a very useful tool to characterize water-in-oil emulsions.21−25 The NMR pulsed field gradient (PFG) pulse sequence provides an ability to measure molecular self-diffusion. When the NMR signal from the droplet phase is selectively detected, the extent of molecular self-diffusion measured within this phase is restricted by the droplet boundary. This can be appropriately analyzed so as to extract emulsion droplet size distributions (DSD).26 Further details are accessible in the following review articles.27,28 The NMR PFG method of determining emulsion droplet sizes can be readily applied to opaque, concentrated emulsions (as is the case for water-in-crude oil emulsions) which preclude the use of optical microscopy and light scattering techniques. In this study, we apply NMR PFG techniques to monitor the increase in emulsion droplet size for a water-in-crude-oil emulsion as a function of CO2 pressure and exposure time. We thus assume that emulsion destabilization occurs via an increase in emulsion droplet size, ultimately leading to breakage of the emulsion into discrete oil and water phases. Accurate monitoring of the emulsion droplet size thus also serves to provide early detection of emulsion destabilization. We further explore this emulsion destabilization process by replacing the CO2 with various other gases of variable solubility in the aqueous and oil phases; knowledge of the gas solubility in both liquid phases is critical to assessing how readily the gas can reach the oil−water interfaces within the emulsion as well as the propensity for (subsequent) bubble formation in the aqueous phase following depressurization. We also consider the treatment of water-in-crude-oil emulsions with either their asphaltene or oil-soluble organic acid contents significantly reduced.

high speed Miccra D-9 homogenizer manufactured by ART Prozess & Laborthechnick, GmbH. Emulsion samples were prepared in 110 g batches and were sheared at 17 800 rpm for 10 min. The emulsion temperature during shearing was monitored and rapidly reached an equilibrium of ∼35 °C, while the pressure was atmospheric. All emulsion samples were stored at room temperature (∼22 °C). 2.2. Crude Oil Modification. Significant literature suggests that naphthenic acids and asphaltenes are two main chemical components that contribute to water-in-crude-oil emulsion stabilization.30−33 In order to elucidate the interplay of CO2 and these components with regard to emulsion stability, asphaltenes and these acids were individually (partially) removed from the crude oil (prior to emulsification) as follows. 2.2.1. Asphaltene Removal. The asphaltene removal was carried out using a technique employed by Pradilla et al.34 The crude oil was diluted with hexane to a volume ratio of 1:30. This was then allowed to rest on the shelf for 24 h to allow for the formation of any asphaltene precipitate (which by definition is insoluble in hexane). The sample was then sequentially filtered using 40, 10, and 0.41 μm pore size filter paper, respectively. The amount of precipitate was gravimetrically determined. The crude oil sample was recovered by removing the hexane using a rotary evaporator. 2.2.2. Acid Removal. Acid removal was achieved using an adsorption process35 via the apparatus shown in Figure 1.

2. METHODOLOGY 2.1. Emulsion Preparation. In this study, we consider two water-in-oil emulsions. The bulk of the work is conducted using a water-in-crude-oil emulsion formulated from a local West Australian crude oil (density 0.81g/mL; viscosity 10 mPa s at 25 °C), known to readily form water-in-oil (w/o) emulsions. The w/o emulsion was formed using 30 wt % deionized (DI) water. To act as a reference, a synthetic emulsion was formulated from paraffin oil (purchased from Sigma-Aldrich with specified density of 0.83 g/mL and viscosity of 110 mPa s at 20 °C). A 20 wt % DI water-in oil emulsion was prepared; this was stabilized by the nonionic surfactant, Span 80 (C24H44O6, 0.5 wt % with respect to the oil phase), which was also purchased from Sigma-Aldrich and which is known to form w/o emulsions.29 Note that it was not possible to form a 30 wt % water-in-paraffin oil emulsion because it inverted to an oil-in-water emulsion. The emulsions were prepared using a

Figure 1. Apparatus set-up used for the acid removal process.

Approximately 40 g of the sorbent Superclean LC-SAX (which contains mainly quaternary amine and counterions) was activated to allow for adsorption of acidic components in the crude oil with 500 mL of 0.1 M NaHCO3/0.1 M Na2CO3 solution, followed by rinse washes with 600 mL of deionized water, 300 mL of methanol, and finally 1000 mL of CH2Cl2. Prior to the acid adsorption process, 40 g (∼50 mL) of the B

DOI: 10.1021/acs.energyfuels.6b00782 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels crude oil was diluted with 500 mL of CH2Cl2. The diluted crude oil was then flowed through the sorbent from above. After the adsorption process had completed, the treated crude oil was stirred in an open beaker under the fume hood for 48 h at ∼25 °C allowing the CH2Cl2 to vaporize. To ensure the removal of any residual CH2Cl2 (boiling point 40 °C), the treated crude oil was then heated to ∼50 °C for 10 min. Gravimetrically, it was confirmed that no significant CH2Cl2 remained. The total acid number (TAN) of the crude oil was measured prior to, and following, this acid removal process using test method ASTM D664.36 The water-in-oil emulsions formed with the oils that had either their asphaltenes or acid partially removed were prepared in the same way as all other emulsions used in this study as detailed in section 2.1. 2.3. CO2 Emulsion Breaking. A 38 mm (i.d.) × 100 mm (l) stainless steel pressure cell, rated to 100 bar, was constructed to enable treatment of the emulsions with high pressure CO2. This is schematically shown in Figure 2. CO2

Figure 3. Schematic of the setup used for the gas solubility measurements.

volume of 110 mL, and was loaded with 50 mL of the crude oil sample. The cell was connected to a high pressure syringe (ISCO) pump loaded with the relevant gas, which was flushed through the void space at atmospheric pressure, prior to elevation of the pressure to 50 bar. Temperature was maintained at 25 ± 0.5 °C using a cooling bath. The crude oil sample was stirred (using a magnetic stirrer) at pressure for 30 min and allowed to equilibrate for a further 24 h. Using the initial (following pressurization) and final gas volume reading on the ISCO pump, we were able to determine the amount of gas that had dissolved in the crude oil. The solubilities of the gases in water were calculated using the volumetric method, with gas density information obtained from the recommended equation of state for the pure fluids37−40 as implemented in the software package REFPROP 9.1.41 2.5. Nuclear Magnetic Resonance Measurements. The main NMR hardware used for this study was a benchtop 1 T permanent magnet featuring a Halbach array and a 5 mm inner diameter rf coil tuned to the 1H resonance of 40 MHz. This magnet arrangement provides a sufficiently homogeneous magnetic field (Water LW50: ∼4 Hz) such that chemical shift resolution of the oil and water NMR signals is readily achieved. A custom-built gradient coil, of maximum strength 1 T/m, was used for all required pulsed field gradient diffusion experiments. The magnet/coil assembly is driven by a Kea2 spectrometer using the software package, Prospa. The entire system was purchased from Magritek, New Zealand. All pulsed field gradient measurements were performed using the stimulated echo pulsed field gradient sequence (STE PFG)27 as shown in Figure 4. Our implementation is detailed in a number of publications,23,24,42,43 the salient features of which are reported here for clarity. With respect to the pulse sequence

Figure 2. CO2 emulsion breaking setup: (a) schematic diagram, and (b) photo of the pressure cell.

pressure was controlled by a pressure regulator with a maximum supply pressure of 55 bar. Initially, 70 mL of the pre-prepared emulsion was loaded into the pressure cell and the cell lid closed. The frit (purchased from Altmann Analytik GmbH & Co. KG, Germany with a 2 μm pore size) at the bottom of the cell prevented the emulsion from draining into the CO2 supply line. The frit also served to generate fine CO2 bubbles (∼2 μm diameter) which were bubbled through the emulsion during treatment for periods varying from 0.5 to 2 h at a flow rate of 50 std cm3/s and pressures from 20 to 50 bar. Measurements were also performed in which the system was allowed to equilibrate to the test pressure and sustained at that pressure with no release of CO2 from the system (hereafter referred to as stationary measurements). After treatment, the cell was periodically depressurized, and a 2 mL emulsion sample extracted and placed in a 5 mm outer diameter NMR tube for droplet sizing analysis as detailed below. This emulsion breaking procedure was repeated for a range of alternative gases, namely, N2, CH4, and N2O. 2.4. Gas Solubility Measurements. Knowledge of the gas solubility in both liquid phases is critical. The solubilities of the four gases considered (CO2, N2, CH4, and N2O) in the crude oil were directly measured. Figure 3 shows the apparatus used for these solubility measurement. The pressure cell has a void

Figure 4. Stimulated echo pulsed field gradient (STE PFG) pulse sequence used for diffusion measurements. C

DOI: 10.1021/acs.energyfuels.6b00782 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

where δjk is the Kronecker delta and αnm is the mth root of the Bessel function:

in Figure 4, the acquired NMR signal, S, can be related to the free self-diffusion of molecules (e.g., water), D, via the Stejskal Tanner equation:44 ⎡ ⎛ S δ ⎞⎤ = exp⎢ −D(γgδ)2 ⎜Δ − ⎟⎥ ⎝ ⎣ S0 3 ⎠⎦

J ′n (α) = 0

The elements of v, describing the reflecting boundary surface without relaxation, are as follows:

(1)

⎧1 for j = 1 vj = δj1 = ⎨ ⎩ 0 for j > 1

Here S0 is the signal measured when g = 0, g is the magnetic field gradient, δ is the duration of the magnetic field gradient applied, γ is the gyromagnetic ratio of the nucleus (1H = 2.68 × 108 T−1 s−1), and Δ is the time interval between two gradient pulses. Hence by varying g while measuring S, D can be extracted by regression of eq 1. The parameters used for STE PFG experiments in this work are δ = 4 μs and Δ = 200 ms with the gradient strength, g, varied from 0 to 1 T/m in 16 increments. For an emulsion, molecular self-diffusion within the droplet phase is restricted by the droplet surface. The extent of restricted diffusion will be determined by the size of the droplets, and it is this physical phenomenon that we exploit to determine the emulsion droplet size distribution in the following manner. Murday and Cotts45 developed a model, based on the assumption of a Gaussian phase distribution (GPD), to describe the NMR signal (S) for restricted diffusion within a spherical droplet of radius a:

⎪ ⎪

v†G(tδ̃ ; γ )exp[ −Λ(Δ − δ)]G*(tδ̃ ; γ )̃ v ̃ S = S(a , g ) = † S0 v exp[−Λ(Δ + δ)]v (9)

Droplet radius (a) can be obtained by regression of eq 9 to relevant experimental NMR data. The BGP method has been demonstrated to provide improved predictions of droplet size over all practical experimental parameter space.50 This method was employed in several previous studies,23,24,42,43 and is the method deployed in this work. Water-in-crude-oil emulsions, however, consist of a range of droplet sizes. Packer and Ress57 employed eq 10 to compute the droplet size distribution, P(a). ∞

b(g ) = ∞

∑ m=1

−

2+e

1 2δ × 2 2 2 2 αm(αma − 2) αmD

−αm2D(Δ− δ)

−αm2DΔ

− 2e (αm2D)2

+e

) (2)

J3/2 (αa) = αaJ5/2 (αa)

(3)

k=1

S = v†(∏ G(tk̃ − tk̃ − 1; γk̃ ))v (4)

where t ̃ = Dt/a is the dimensionless time and γ̃ = −cγga /D is the dimensionless magnetic field gradient and 2

G(t ;̃ γ )̃ = exp[( −Λ + i(γ ·̃ z)̃ Z)t ]̃

∞

∫0 a3P(a)da

(10)

3. RESULTS AND DISCUSSION 3.1. Effect of CO2 Pressure and Exposure Duration on Emulsion Stability. The effect of CO2 on the water-in-crude oil emulsion stability was investigated at pressures from 20 to 50 bar. Following the 2 h CO2 treatment, NMR measurements of the emulsion droplet size distribution (as detailed above) were performed periodically following depressurization. The emulsion droplet size distributions measured immediately following treatment are shown in Figure 5. As the pressure of the CO2 during treatment is increased, the droplet size distribution clearly shifts to larger values. The mean emulsion droplet size from the distributions shown in Figure 5 was determined and is reported in Figure 6 as a function of pressure and time following emulsion formation. The stability of the untreated emulsion is clearly evident. The increase in emulsion water droplet size with treatment pressure was distinct and sustained. The effect is most evident at a pressure of 50 bar (where the droplet sizes approximately triple), minimal at 30 bar, and negligible at 20 bar. While previous literature17 has focused on emulsion destabilization at supercritical conditions, we demonstrate here for the first time, to the best of our knowledge, that subcritical CO2 can have a significant effect on water-in-crude-oil emulsion stability. It also highlights the strength of the NMR measurements as a screening technique that can readily detect micronscale changes in mean droplet size. For clarity and ease of

and Jn is the nth order Bessel function. Linse and Soderman46 also proposed the use of a short gradient pulse (SGP) model which assumes the gradient pulse duration is infinitely short, δ = 0, and the diffusion that occurs during this period can hence be neglected. Both GPD and SGP have their drawbacks;47−49 Lingwood et al.50 implemented the block gradient pulse (BGP) method to overcome these limitations based on general gradient waveforms51−53 to quantify the signal attenuation. The block gradient pulse method solves the relevant Block Torrey equation54 using eigenfunction expansions.48 The signal intensity, when the pulse sequence is divided into Ne intervals of constant gradient, is expressed as

Ne

∫0 a3P(a)S(a , g )da

Here b(g) is the normalized NMR signal as a function of g. This is, however, an ill-conditioned inversion problem. Hollingsworth and Johns26 showed how Tikonov regularization techniques can be used to invert eq 10, where the required smoothing parameter is determined using the generalized cross validation (GCV) method. This regularization technique has been extensively validated and used in previous studies.23,24,42,43

−αm2D(Δ+ δ)

where αm are the positive roots of the following function:

3

(5)

where matrix Z correlates the equilibrium mode to the higher eigenmodes for a sphere.49,55,56 The elements of matrix Λ are computed using 2 Λ = αnm δjk

(8)

The resultant BGP signal attenuation function is as follows:

S = S(a , g ) S0 = exp( −2γ 2g 2

(7)

(6) D

DOI: 10.1021/acs.energyfuels.6b00782 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

Given the effectiveness of the 50 bar CO2 treatment, we proceeded to explore the effect of treatment duration at this pressure. Figure 7 shows the effect of CO2 treatment durations from 0.5 to 2 h (separate experiments were performed for each treatment duration) on the emulsion droplet size distribution. Figure 7a shows the measurements corresponding to the flow of the CO2 through the emulsion sample at 50 mL/s (STP) while Figure 7b shows the measurements when the sample was simply pressurized to 50 bar in a closed system using CO2. In both cases it is evident that treatment duration had a pronounced effect on emulsion stability. However, at intermediate treatment times the effect on emulsion stability is more pronounced in Figure 7a, which shows the flowing CO2 scenario, indicating that mass transfer limitations are reduced in this case. These could either be dissolution into the oil phase or across the interface and into the aqueous droplet phase. However, no difference is evident for a 2 h duration between the two treatment scenarios, which is indicative that mass transfer limitations are not affecting the overall emulsion destabilization extent at these conditions. Treatment durations of 2 h at 50 bar were thus applied in all subsequent experiments. 3.2. Effect of Various Gases on Emulsion Stability. The emulsion treatment process was repeated at 50 bar for 2 h with three alternative gases to CO2: N2, CH4, and N2O. Table 1

Figure 5. Droplet size distributions (DSDs) for water-in-crude oil emulsions following treatment with CO2 at various pressures.

Table 1. Solubilities in Crude Oil (Measured) and Water58 gas CO2 N2 CH4 N2O

solubility in water (50 bar, 25 °C) 1.74 0.03 0.07 1.20

mol/L mol/L mol/L mol/L

solubility in crude oil (50 bar, 25 °C) 3.80 0.73 1.09 7.60

mol/L mol/L mol/L mol/L

reports their respective solubilities in the crude oil, as well as in water. Figure 8 shows emulsion stability following treatment with the respective gases. The N2 and CH4 clearly have no effect on emulsion destabilization, which is consistent with their low solubilities in both water and the crude oil. In contrast, the N2O behaved similarly to CO2 in terms of its effect on emulsion stability. With reference to Table 1, N2O was twice as soluble in the crude oil as CO2 but about 45% less soluble in the aqueous phase. CO2 and N2O were both significantly more

Figure 6. Mean droplet sizes for water-in-crude-oil emulsions with and without CO2 treatment over 5 h. The gray region corresponds to the treatment period (if one occurred) with CO2 at the pressure indicated. At other times, the pressure applied to the emulsion was atmospheric.

comparison, all subsequent emulsion measurement data are presented in terms of the mean droplet size.

Figure 7. Measurements of CO2 treated and untreated water-in-crude-oil emulsion mean droplet sizes over 5 h under (a) continuous CO2 flow treatment and (b) stationary conditions (as detailed in the text). The gray regions correspond to the different CO2 treatment durations. All measurement were performed at 50 bar and 25 °C. E

DOI: 10.1021/acs.energyfuels.6b00782 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

Figure 8. Measurements of (a) CO2, (b) N2O, (c) CH4, (d) N2 treated and untreated crude oil emulsion mean droplet sizes over 5 h. The gray regions correspond to the CO2 treatment duration.

soluble in the aqueous phase relative to N2 and CH4; however, they are only moderately more soluble in the crude oil. This suggest that solubility in the aqueous phase is more relevant to emulsion destabilization; this however is somewhat speculative and requires more detailed investigation. It must also be acknowledged that solubility is not the only contribution to destabilization; the chemical interactions of the gas with the crude oil are also potentially relevant. 3.3. Effect of Liquid Composition on Emulsion Stability. We replaced the crude oil with paraffin oil and formulated a stable emulsion with 20 wt % water using 0.5 wt % Span80 (with regard to the oil phase), as detailed above. Consistent with the data in Figure 8, this was then subjected to a treatment of 50 bar CO2 for 2 h followed by depressurization and subsequent periodic droplet sizing. The resultant evolution in the mean droplet size is summarized in Figure 9. The droplets formed in this emulsion were approximately 50% larger than those formed in the water-in-crude-oil emulsion. The effect of the CO2 treatment was, however, negligible. This result is inconsistent with the mechanism proposed in theory 1 (as detailed above in section 1) in which CO2 dissolves in the aqueous droplet phase, ruptures through the droplet interface, and destabilizes the emulsion. Hence the lack of emulsion

Figure 9. Measurements of CO2 treated and untreated water-inparaffin-oil emulsions’ mean droplet sizes over 5 h. The gray region corresponds to the treatment duration.

destabilization observed provides evidence for the mechanism proposed in theory 2, in which the CO2 causes asphaltenes to F

DOI: 10.1021/acs.energyfuels.6b00782 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels

indicating that CO2 enhancement of asphaltene precipitation in the crude-oil emulsion was most likely the destabilization mechanism. The effect was replicated when CO2 was replaced by N2O; it is worth noting that both gases have significant solubilities in both the crude oil and water at the treatment conditions. Partial removal of the asphaltenes prior to the emulsion formation resulted in the CO2 treatment being significantly more effective. Partial removal of napthanic acid content resulted in a significantly more unstable initial emulsion but did not enhance the subsequent destabilization effect of the CO2 treatment.

precipitate out of solution resulting in emulsion destabilization. The results in Figure 8, however, suggest that this is not unique to CO2 and that N2O also interacts with the asphaltenes. To explore this further, the same treatment process was applied to water-in-crude oil emulsions formed from oil phases where either the asphaltene or napthanic (oil-soluble) acid content was significantly reduced, as detailed above. Results for the sample with acid removed are shown in Figure 10a. The

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Support of the Australian Research Council via Grant DP130101461 is gratefully acknowledged. Special thanks go to UWA undergraduate students, Douglas Vehlo and Nikita Shah, for their assistance in conducting NMR measurements.

Figure 10. Measurements of CO2 treated and untreated (a) acidremoved (the total acid number of this sample was reduced by 50%) and (b) asphaltenes-removed (0.31 wt % was removed) crude oil emulsion mean droplet sizes over 5 h. The gray regions correspond to the treatment duration.

■

total acid number of this sample was reduced by 50% from 0.2 to 0.1 mg KOH/g. This clearly affected the stability of the initial emulsion, which featured approximately 50% larger droplets that were slightly less stable over time. However, it did not accentuate the subsequent destabilization effect of the CO2: the final mean droplet size was comparable to that in Figure 8a. With respect to the sample with asphaltene reduction, the results are shown in Figure 10b: the initial emulsion appears unaffected by the removal of the asphaltene (0.31 wt % was removed, as determined gravimetrically). However, in this case the CO2 was significantly more effective at destabilizing the emulsion in contrast with the effect of CO2 on the original crude oil emulsion as shown in Figure 8a, increasing the mean droplet size by almost an order of magnitude and ultimately breaking the emulsion into its constituent phases. This suggests that the CO2 to asphaltenes ratio has reached a critical level, above which the emulsions will destabilize rapidly. We speculate that in this case, consistent with theory 2, CO2 precipitated out the remnant asphaltenes and thus rendered the emulsion considerably less stable. To explore the relative interplay between theories 1 and 2 with regard to the effect of CO2 on emulsion (water-in-crudeoil) stability further, we are currently constructing a carbonicacid-compatible NMR cell in which emulsion droplet sizing can be performed at pressure, prior to depressurization. This should provide conclusive proof as to whether depressurization and CO2 bubble formation significantly affect emulsion stability. We will apply this measurement to a range of crude oils featuring a range of asphaltene and acid contents.

REFERENCES

(1) Ferro, B. D.; Smith, M. Global Onshore and Offshore Water Production; Douglas-Westwood Ltd, 2007. (2) Arnold, R.; Burnett, D. B.; Feeley, T. J.; Galbrun, M.; Hightower, M.; Jiang, Z.; Khan, M.; Lavery, M.; Luffey, F.; Verbeek, P. Managing Water-From Waste to Resource 2004. (3) Alboudwarej, H.; Muhammad, M.; Shahraki, A. K.; Dubey, S.; Vreenegoor, L.; Saleh, J. M. Rheology of Heavy-Oil Emulsions. SPE Production & Operations 2007, 22 (3), 285−293. (4) Been, J.; Crozier, B.; Mosher, M.; Mosher, W.; Tsaprailis, H. Internal Corrosion of Crude Oil Pipelines Research; Alberta Innovates Technology Futures, 2012. (5) Kokal, S. L. Crude Oil Emulsions. In Petroleum Engineering Handbook; Fanchi, J. R., Ed.; Society of Petroleum Engineers: Texas, 2006; Vol. 1, pp 533−570. (6) Salager, J. L. Surfactants in Aqueous Solutions; Universidad De Los Andes: Mérida-Venezuela, 1994. (7) Hill, M.; Parker, S. P. MacGraw-Hill Dictionary of Scientific & Technical Terms, 6th ed.; The McGraw-Hill Companies, Inc., 2003. (8) Abdel-Aal, H. K.; Aggour, M.; Fahim, M. A. Petroleum and Gas Field Processing; Marcel Dekker, Inc., 2003. (9) Binks, B. P.; Horozov, T. S. Colloidal Particles at Liquid Interfaces. Cambridge University Press: Cambridge, U.K., 2006. (10) Warren, K. W. Emulsion Treating. In Petroleum Engineering Handbook: Facilities and Construction Engineering; Arnold, K. E., Ed.; Society of Petroleum Engineers, 2006; Vol. 3, pp 63−122. (11) Satter, A.; Iqbal, G. M.; Buchwalter, J. L. Practical Enhanced Reservoir Engineering: Assisted with Simulation Software; PennWell Corporation: U.S., 2008. (12) Silset, A. Emulsions of Heavy Crude Oils. Characterization, Stabilization, Destabilization and Produced Water Quality; Norweigian University of Science and Technology, 2008. (13) Sjoblom, J.; Kallevik, H.; Westvik, A.; Auflem, I. H. Process for Separation of Oil. Water and Gas in Separator by Breaking Water-in-oil Emulsions 2002. (14) Little, C. T. Removal of Petroleum Hydrocarbons from Oil-Field brines with Carbon Dioxide; The University of Texas, Austin, 1990. (15) Burke, N. C.; Curtice, S.; Little, C. T.; Seibert, A. F. Removal of Petroleum Hydrocarbons from Oil-Field Brines with Carbon Dioxide. Society of Petroleum Engineers 1991, 21046, 467−465.

■

CONCLUSIONS Treatment of water-in-crude-oil emulsions with CO2 at 50 bar for a duration of 2 h resulted in significant emulsion destabilization as determined by NMR measurements of droplet size distributions. This was not replicated for a model emulsion (water-in-paraffin oil featuring a nonionic surfactant) G

DOI: 10.1021/acs.energyfuels.6b00782 Energy Fuels XXXX, XXX, XXX−XXX

Article

Energy & Fuels (16) Auflem, I. H.; Kallevik, H.; Westvik, A.; Sjoblom, J. Influence of Pressure and Solvency on the Separation of Water-in-Crude Oil Emulsions from the North Sea. J. Pet. Sci. Eng. 2001, 31 (1), 1−12. (17) Zaki, N. N.; Carbonell, R. G.; Kilpatrick, P. K. A Novel Process for Demulsification of Water-in-Crude Oil Emulsions by Dense Carbon Dioxide. Ind. Eng. Chem. Res. 2003, 42 (25), 6661−6672. (18) Akbarzadeh, K.; Huo, Z.; Broze, G. Flow-through Tests Advance Researchers’ Understanding of Asphaltene Deposition. World Oil 2010. (19) Srivastava, R. K.; Huang, S. S.; Dong, M. Asphaltene Deposition During CO2 Flooding. SPE Prod. Facil. 1999, 14 (4), 235−245. (20) Zanganeh, P.; Ayatollahi, S.; Alamdari, A.; Zolghadr, A.; Dashti, H.; Kord, S. Asphaltene Deposition During CO2 Injection and Pressure Depletion: A Visual Study. Energy Fuels 2012, 26, 1412− 1419. (21) Hindmarsh, J. P.; Hollingsworth, K. G.; Wilson, D. I.; Johns, M. L. An NMR Study of the Freezing of Emulsion-Containing Drops. J. Colloid Interface Sci. 2004, 275 (1), 165−171. (22) Voda, M. A.; van Duynhoven, J. Characterization of Food Emulsions by PFG NMR. Trends Food Sci. Technol. 2009, 20 (11−12), 533−543. (23) Ling, N. N. A.; Haber, A.; May, E. F.; Fridjonsson, E.; Johns, M. L. NMR Studies of Emulsion Microstructure Approaching the Phase Inversion Point. Colloids Surf., A 2014, 462, 244−251. (24) Ling, N. N. A.; Haber, A.; Fridjonsson, E. O.; May, E. F.; Johns, M. L. Shear-induced Emulsion Droplet Diffusion using NMR. J. Colloid Interface Sci. 2016, 464, 229−237. (25) Lonnqvist, I.; Khan, A.; Soderman, O. Characterization of Emulsions by NMR Methods. J. Colloid Interface Sci. 1991, 144 (2), 401−411. (26) Hollingsworth, K. G.; Johns, M. L. Measurement of Emulsion Droplet Sizes Using PFG NMR and Regularization Methods. J. Colloid Interface Sci. 2003, 258, 383−389. (27) Johns, M. L.; Hollingsworth, K. G. Characterisation of Emulsions Systems using NMR and MRI. Prog. Nucl. Magn. Reson. Spectrosc. 2007, 50, 51−70. (28) Johns, M. L. NMR Studies of Emulsions. Curr. Opin. Colloid Interface Sci. 2009, 14, 178−183. (29) Morales, C. J.; Riebel, U.; Guzmán, N. M.; Guerra, M. Formulation of Water-in-Paraffin Emulsions. Latin American Applied Research 2011, 41 (2), 105. (30) Bestougeff, M. A.; Byramjee, R. J. Asphaltenes and Asphalts, 1; Elsevier Science: Amsterdam, 1994. (31) Gafonova, O. V.; Yarranton, H. W. The Stabilization of Waterin-Hydrocarbon Emulsions by Asphaltenes and Resins. J. Colloid Interface Sci. 2001, 241 (2), 469. (32) Chaverot, P.; Cagna, A.; Glita, S.; Rondelez, F. Interfacial Tension of Bitumen-Water Interfaces. Part 1: Influence of Endogenous Surfactants at Acidic pH. Energy Fuels 2008, 22 (2), 790−798. (33) Simon, S.; Nordgård, E.; Bruheim, P.; Sjöblom, J. Determination of C80 Tetra-acid Content in Calcium Napthenate Deposits. Journal of Chromatography A 2008, 1200 (2), 136−143. (34) Pradilla, D.; Simon, S.; Sjoblom, J. Mixed Interfaces of Asphaltenes and Model Demulsifiers Part I: Adsorption and Desorption of Single Components. Colloids Surf., A 2015, 466, 45−56. (35) Gawel, B.; Bandyopadhyay, S.; Øye, G.; Lesaint, C. The Role of Physicochemical and Interfacial Properties on the Binary Coalescence of Crude Oil Drops in Synthetic produced Water. Energy Fuels 2014, 512. (36) ASTM. Standard Test Method for Acid Number of Petroleum Products by Potentiometric Titration; 2011; Vol. D664-11a. (37) Span, W.; Wagner, W. A New Equation of State for Carbon Dioxide Covering the Fluid Region from the Triple-Point Temperature to 1100 K at Pressures up to 800 MPa. J. Phys. Chem. Ref. Data 1996, 25 (6), 1509−1596. (38) Lemmon, E. W.; Span, W. Short Fundamental Equations of State for 20 Industrial Fluids. J. Chem. Eng. Data 2006, 51, 785−850. (39) Span, W.; Lemmon, E. W.; Jacobsen, R. T.; Wagner, W.; Yokozeki, A. A Reference Equation of State for the Thermodynamic

Properties of Nitrogen for Temperatures from 63.151 to 1000 K and Pressures to 2200 MPa. J. Phys. Chem. Ref. Data 2000, 29 (6), 1361− 1433. (40) Setzmann, U.; Wagner, W. A New Equation of State and Tables of Thermodynamic Properties for Methane Covering the Range from the Melting Line to 625 K at Pressures up to 1000 MPa. J. Phys. Chem. Ref. Data 1991, 20 (6), 1061−1151. (41) Lemmon, E. W.; Huber, M. L.; McLinden, M. O. NIST Standard Reference Database 23: Reference Fluid Thermodynamic and Transport Properties-REFPROP version 9.1; National Institute of Standards and Technology: Gaithersburg, 2013. (42) Fridjonsson, E.; Graham, B.; Akhfash, M.; May, E.; Johns, M. L. Optimized Droplet Sizing of Water-in-Crude Oil Emulsions using Nuclear Magnetic Resonance. Energy Fuels 2014, 28, 1756−1764. (43) Haber, A.; Akhfash, M.; Loh, C. K.; Aman, Z. M.; Fridjonsson, E. O.; May, E. F.; Johns, M. L. Hydrate Shell Growth using NMR. Langmuir 2015, 31, 8786−8794. (44) Tanner, J. E.; Stejskal, E. O. Restricted Diffusion of Protonsin Colloid Systems by the pulsed field gradient method. J. Chem. Phys. 1968, 49, 1768−1777. (45) Murday, J. S.; Cotts, J. M. Self-diffusion Coefficient of Liquid Lithium. J. Chem. Phys. 1968, 48, 4938−4945. (46) Linse, P.; Soderman, O. The Validity of the Short-GradientPulse Approximation in NMR Studies of Restricted Diffusion. Simulations of Molecules Diffusing between Planes, in Cylinders and Spheres. J. Magn. Reson., Ser. A 1995, 116 (1), 77−86. (47) Barzykin, A. V. Exact Solution of the Torrey-Bloch Equation for a Spin Echo in Restrcited Geometries. Phys. Rev. B: Condens. Matter Mater. Phys. 1998, 58 (21), 14171−14174. (48) Barzykin, A. V. Theory of Spin Echo in Restricted Geometries under a Step-wise Gradient Pulse Sequence. J. Magn. Reson. 1999, 139 (2), 342−353. (49) Grebenkov, D. S. NMR Survey of Reflected Brownian Motion. Rev. Mod. Phys. 2007, 79, 1077−1137. (50) Lingwood, I. A.; Chandrasekera, T. C.; Kolz, J.; Fridjonsson, E. O.; Johns, M. L. Emulsion Droplet Sizing using Low-field NMR with Chemical Shift Resolution and the Block Gradient Pulse Method. J. Magn. Reson. 2012, 214, 281−288. (51) Caprihan, A.; Wang, L. Z.; Fukushima, E. A Multiple-narrowpulse Approximation for Restricted Diffusion in a Time-varying Field Gradient. J. Magn. Reson., Ser. A 1996, 118 (1), 94−102. (52) Callaghan, P. T. A Simple Matrix Formalism for Spin Echo Analysis under Generalized gradient Waveforms for Spherical Pores with Relaxivity and Interconnections. J. Magn. Reson. 1997, 129 (1), 74−84. (53) Ryland, B. N.; Callaghan, P. T. Spin Echo Analysis of Restricted Diffusion Under Generalized Gradient Waveforms for Spherical Pores with Relaxivity and Interconnections. Isr. J. Chem. 2003, 43 (1−2), 1− 7. (54) Torrey, H. C. Bloch Equations with Diffusion Terms. Phys. Rev. 1956, 104, 563. (55) Grebenkov, D. S. Laplcaian Eigenfunctions in NMR I. A Numerical Tool. Concepts Magn. Reson., Part A 2008, 32A (4), 277− 301. (56) Grebenkov, D. S. Laplacian Eigenfunctions in NMR II Theoretical Advances. Concepts Magn. Reson., Part A 2009, 34A (5), 264−296. (57) Packer, K. J.; Rees, C. Pulsed NMR Studies of Restricted Diffusion. I. Droplet Size Distributions in Emulsions. J. Colloid Interface Sci. 1972, 40 (2), 206. (58) Sander, R. Compilation of Henry’s Law Constants (Version 4.0) for Water as Solvent. Atmos. Chem. Phys. 2015, 15, 4399−4981.

H

DOI: 10.1021/acs.energyfuels.6b00782 Energy Fuels XXXX, XXX, XXX−XXX