Influence of thermobaric conditions on size distribution of colloidal gas aphrons

Accepted Manuscript Title: Influence of Thermobaric Conditions on Size Distribution of Colloidal Gas Aphrons Author: Oleg Zozulya Vera Pletneva PII: DOI: Reference:

S0927-7757(15)00434-3 http://dx.doi.org/doi:10.1016/j.colsurfa.2015.05.039 COLSUA 19959

To appear in:

Colloids and Surfaces A: Physicochem. Eng. Aspects

Received date: Revised date: Accepted date:

11-2-2015 9-5-2015 11-5-2015

Please cite this article as: Oleg Zozulya, Vera Pletneva, Influence of Thermobaric Conditions on Size Distribution of Colloidal Gas Aphrons, Colloids and Surfaces A: Physicochemical and Engineering Aspects http://dx.doi.org/10.1016/j.colsurfa.2015.05.039 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Title: Influence of Thermobaric Conditions on Size Distribution of Colloidal Gas Aphrons Author(s): Oleg Zozulya (Schlumberger Moscow Research), Vera Pletneva (Moscow State University)

Corresponding author: Oleg Zozulya e-mail: [email protected] mobile: +79168081801 phone: +74959358200 ext.6024019

Graphical abstract

Highlights    

preparation of monodisperse CGA stable at thermobaric conditions of a reservoir high-pressure instrumentation for CGA behavior monitoring at reservoir conditions improved Circular Hough Transform for detection of >80% of bubbles present on image key parameters affecting CGA size distribution during preparation and coarsening

Abstract A new technique for the preparation of monodisperse colloidal gas aphrons (CGA) stable at reservoir conditions is developed. The behaviour of such aphrons under compressiondecompression protocols with pressure in the range 1÷500 bar and pressure release rate in the range 50÷3000 bar/min is studied with optical microscopy and densimetry. An image processing algorithm is developed for counting the aphrons and measuring their size distribution. It is shown that the degree of polydispersity of the aphron system is independent of pressure. The dependence of the mean aphron size versus time follows Lifschitz-Slyozov law which is characteristic of wet foam coarsening. The monodisperse CGAs produced can be successfully used as pressure-temperature controlled systems for various oilfield applications Keywords: Colloidal Gas Aphron, Xanthane, High-pressure Microscopy, Circular Hough Transform, Bubble nucleation, Ostwald ripening, Lifshitz-Slyozov-Wagner theory

Introduction Preparation of stable microbubble mixtures is of growing interest in various oilfield frontiers: drilling, completions, water flood monitoring, hydraulic fracturing etc. The main properties of bubble mixtures that attract much of the attention are their low density, effective gas encapsulation, and high yield stress, shear thinning rheology. The main complication in the development of the successful oilfield applications lies in the production of bubble mixtures able to survive reservoir conditions. The stability of micron-size bubbles comprises the problem that has been thoroughly studied in the field of medical ultrasound diagnostics where stabilized microbubbles are widely used as intravenously injected contrast agents [1]. The long life-time of such microbubbles is achieved as a result of four effects: the low diffusivity and the low solubility of the suitable gas in water solutions, the vanishing surface tension of the gas-liquid interface and the hardening (increase of elasticity) of the bubble shell. If one tries to transfer these stabilization techniques to reservoir conditions he would inevitably arrive at two possibilities: apply hard shell bubbles like hollow glass spheres [2] or generate soft shell bubbles like colloidal gas aphrons (CGA) which are capable of rearranging (reinforcing) their shells in response to pressure increase [3]. The former is obviously the most expensive case, moreover its disadvantage lies in the fact that the hard particles remaining intact during production operations can get accumulated, block pores and thus decrease permeability near injection wells. The latter is more appropriate case for the oilfield applications as wellbore deblocking is easily achieved on pressure release [3].

CGA structure At ambient conditions CGA based fluids consist of microbubbles (10÷100 m), the cores of which are composed of a gas surrounded by a thick multilayer shell [4]. The shell is formed of an inner surfactant film enveloped by a viscous water layer,

Fig.1. CGA multilayer structure.

which is in turn covered by a double layer of surfactants that provides rigidity and low permeability to the structure while imparting to it some hydrophilic character (see Fig.1). The polar heads of the molecules that make up the outermost surfactant layer are oriented into the aqueous carrying fluid, thus making the structure hydrophilic and dispersible in it. This outermost surfactant layer also imparts an anionic charge to the aphron. Thus, aphrons tend to have little affinity for each other or for anionic mineral surfaces. These features ensure the successful use of CGA in drilling operations: by effective sealing of the wellbore CGAs prevent mud fluid losses and formation damage. On Fig.1 you can see as well a microscope image of a 1

gas aphron taken at ambient conditions. The noticeable eccentricity of the inner and the outer shell supports the outlined double-layer structure of the aphron as it can’t be attributed to an optical artefact. In the work the notion of the CGA behaviour at reservoir conditions is acquired through the monitoring of the aphron density and the bubble size distribution (BSD) under variable pressure protocols.

CGA Preparation Procedure The CGA based fluids used for the studies are prepared with the anionic surfactant sodium dodecyl sulfate (SDS) as an aphronizer. As a component of CGA, it is used to reduce the surface tension in order to preserve the gas aphron as it is formed and build the multilayer bubble wall. A polymer component of CGA acts as a viscosifier as well as a stabilizer. It improves the bubble stability by decreasing both the fluid drainage rate and the gas diffusion through the bubble interface. A set of test experiments is performed to find out what biopolymer among those actively used in oilfield applications is the best candidate for the preparation of the gas aphrons stable at reservoir conditions. Among the tested biopolymers are such polysaccharides as potassium alginate, guar, and xanthane. Imposing the same pressure protocol we varied polymer and surfactant concentration and determined the highest pressure value up to which the corresponding CGA survive. Our tests show that xanthane is the best choice in this respect. Moreover xanthane is the low shear rate viscosifier, which is used to control invasion of drilling fluid and to clean borehole from cutting suspension. As CGA based drilling fluid reduces deep fluid invasion both rheologically and mechanically it displays more attractive behaviour in comparison with pure xanthane solution. Our rheological experiments show that the gas aphrons increase the viscosity by 56 times and such parameters of Bingham model as yield stress and flow behaviour index by 1.5-2.5 times. Usually CGAs are generated in a viscous polymer/surfactant solution with a high-speed blade mixer. Decompression voids emerging in such turbulent rotational flow give rise to the formation of bubbles of various sizes and as a result the final BSD of the produced CGA is broad. Broad BSD is obviously a drawback if one attempts to produce CGA Fig.2. CGA foam generation

based fluid with pressure-controllable mean size. Indeed, the aphron colloid with broad BSD is vastly

subjected to inter-bubble gas diffusion as it contains a lot of bubbles that eventually become 2

subcritical and dissolve. To produce narrow bubble SDs mechanical homogenization is used as a first step needed to quickly entrap gas into the base polymer/surfactant solution, then the decompression approach is applied, i.e. on the second step the aphron foam is squeezed till all the amount of the entrapped gas get dissolved in the solution, and finally the gas-saturated solution is exposed to the abrupt pressure release with controllable pressure release rate. It results in gas nucleation and in the formation of the monodisperse CGA based fluid. Further the preparation procedure for the most stable CGA based on xanthane gum is presented in detail. At first stock aqueous solutions of 15g/L xanthane, 200 g/L SDS and 15g/L sodium stearate (NaSt) are prepared. The first two solutions are obtained under stirring at room temperature until the complete dissolution of the chemicals. Sodium stearate 15g/L solution is made by mixing stearic acid and 2 wt.% NaOH solution in a heated environment (85ºC) for 2 hours under slow stirring. At the end of heating NaSt solution is quickly added to xanthane-SDS solution which is formed by mixing the desired ratio (2:1) of xanthane and SDS stock solutions. Then the mixture is vigorously homogenized for 7 minutes at 12000 rpm with a mixer specially constructed in the house for the aphron preparation (see Fig.2). The mixer and the vessel containing the solution are enclosed within a thin transparent plastic shell which is connected to a gas line providing nitrogen at the rate of 250 cm3/min. The shell is used to maintain single gas atmosphere during the preparation of the aphron foam and to ensure the reproducibility of the results (thus both the gas diffusion and the gas dissolution are maintained at the same values despite the case where the gas is air). On this stage the output sample is a coarsely dyspersated bubble foam with nitrogen as an encapsulated gas, it has wide BSD and the aphrons are of nonspherical highly-deformed shape. The mixture volume increases by 5-6 times as a result of the gas entrapment during the mixer homogenization (final nitrogen content of the foam is 0.830

0.85). This preparation stage is performed at ambient conditions (1 bar, 22 C). On the second stage the foam is quickly squeezed to the nitrogen-saturated solution at 350÷450 bar. The actual value is selected on the basis of the gas saturation pressure which is dependent on the solution 0 temperature (e.g. the gas saturation pressure is 320 bar at 22 C). The solution is left pressurized for 3 hours in order to secure the thermal equilibrium. Finally, on the third stage, the solution is decompressed at a fixed pressure release rate to the prescribed pressure value below the gas nucleation threshold. As there are two possibilities by which the dissolved gas may find its way out of the solution, i.e. the growth of the already nucleated bubbles and the nucleation of new ones, the shorter is the nucleation stage (the higher the pressure release rate) the more monodisperse is the resultant CGA based fluid.

High-pressure Optical Microscopy The BSD of the prepared CGA based fluids is characterized by an optical microscopy. A thermostabilized high-pressure optical cell is specially designed for use with Carl Zeiss inverted 3

Fig.3. Inverted microscope setup for CGA SD analysis. HP optical cell is on microscope stage. microscope (Axio Observer.A1m). The main problem was a short focal length of the microscope lens system which prevented from using thick optical windows needed for HP experiments. Our design (see Fig.3) based on fused quartz windows Ø7.5 x 3.5 mm manufactured with deep polishing procedure (for HP-applications) is capable to withstand pressure up to 500 bar. Due to high turbidity of the aphron samples an auxilary light source is used to illuminate the sample from backside via an optical fiber cable. The microscope is equipped with a 12Mpx digital camera which enables to acquire color images with exposure interval of 100 ms. The acquired CGA images are processed in order to determine the radius and the position of the aphrons.

4

The pressure variations in the optical cell are maintained with the help of the inhouse made thermostabilized PVT-cell (see Fig.3) which secure the required foam volume changes (up to 7 times) for the pressure range of studies: 1÷500 bar. The PVT-cell assembly consists of a cylinder with a moveable piston in it which separates the pumping volume from the sample volume. The thermal stabilization is ensured by enclosing the cylinder into a jacket with circulating agent supplied from Julabo thermostat. The sample volume in the PVT-cell is controlled with a position digitizer connected to the cell piston via a rigid rod (precision is 0.01 mm or 20 μl). In turn the pumping of the PVT-cell is performed with ISCO pump system, model 260D from Teledyne Inc., capable to provide flow rates up to 100 cm3/min and support constant pressure gradient in time up to 3200 bar/min. Temperature and pressure are logged via sensors integrated into the both cells. As the aphron foams are very viscous the sample loading is performed in two steps. First we disconnect the PVT-cell from the optical cell, load the PVTcell with the aphron foam and run pump at maximal flow rate in order to force the remaining air out. As we start geting continuous phase of the apron foam from the outlet line of the PVT-cell we stop the pumping and connect it to the optical cell and repeat the procedure. Finally we disconnect the cells from atmosphere and apply the prescribed pressure and temperature protocols.

Image Processing Algorithm Locating geometric figures is a common problem in image analysis and robot vision. The class of algorithms known as the Hough transform has long been a standard method for solving this problem due to their robustness for partially occluded objects and to the presence of noise [5]. The original Hough transform maps an image into an abstract parameter space via a voting process. Each significant pixel in the input image is analyzed in order to determine whether it is a part of some instance from an indexed family of figures. Each pixel can therefore be considered to generate a “vote” for each of the plausible target objects. In practice, votes are accumulated in an abstract parameter space known as the Hough space, which has dimension equal to the number of parameters required to uniquely describe the figure in question. A Hough transform intended to find arbitrary circles has a three-dimensional Hough space, with three parameters corresponding to the 2D coordinates of the centre and the radius of the circle, respectively. The Hough space is discretized into an appropriate number of cells (a compromise between resolution and memory needs), each of which acts as an accumulator of votes. When the transform is complete the Hough space is searched for peaks. The location of each peak corresponds to a geometric object that is present in the image. Circular Hough transform (CHT) is a very robust algorithm however 3D CHT computations are time and memory-consuming. Several modifications are used to reduce the time and storage demands. The most popular are the algorithms that reduce the dimensionality of the Hough 5

space and the statistical ones that do not completely sample the image. The authors apply the former approach and use a 2D CHT implementation from OpenCV library as a starting point [6]. In the implementation the 2D accumulator is used for detecting the circle centres. First, Canny edge detection if performed and each nonzero point on the edge image is analyzed in order to determine the gradient (normal) direction. The accumulator gets incremented for all the points of the normal that are located within the considered range of radii from the taken edge point. On completion of the edge point processing all the potential circle centres sorted in descending order in accumulator values are then being analysed in order to determine the most supported (with maximal number of nonzero points on edge image) value of the radius. Such two-stage processing reduces the dimensionality of the Hough space and the computational time. We made several improvements to the implementation that enables to detect about 70-80% of the bubbles present on a microscope image (see Fig.4).

Fig.4. An example of aphron detection with implemented 2D CHT (1/9 area fragment of microscope image is shown). Initial pressure is 400 bar, final pressure is 10 bar, pressure decrease rate is 1600 bar/min, images are taken after lapse of: 1 min (1), 5 min (2), 40 min (3), and 118 min (4) since the end of decompression stage.

Fig.5. 1/9 area fragments of the microscope images taken at the end of decompression stage, pressure release rate is 800 bar/min, temperature is 400C, initial pressure is 450 bars, final pressure is: 320 bar (1), 270 bar (2), 210 bar (3), 160 bar (4), 80 bar (5), 40 bar (6), 20 bar (7), 10 bar (8), 5 bar (9), and 2.2 bar (10). For all the images the aphron size histogram (aphron SD), bubble mean radius , standard deviation = ( - 2)1/2and the total bubble volume are calculated for each stabilized value of pressure and temperature. As well , /, and the instantaneous radius of 6

individual bubbles R k are traced in time in order to validate the process of inter-bubble gas diffusion.

Results and Discussion The most important issues that arise when one thinks over possible applications of CGAs in oilfield are the CGA stability in time and the possibility to control the aphron SD by pressure and temperature variation. Fig.5 represents the microscope images that illustrate the dependence of the aphron SD on the final pressure at the decompression stage (gas supersaturation). As it is seen from the figure the gas aphrons are stable at the pressure of 320 bar and have the radius of ~2 m. One should bear in mind that the ideal gas law may be used to estimate the contraction/expansion of the bubble foam only for pressure values far from the saturation pressure. Indeed the figure enables to estimate the ratio of the aphron size at close to normal pressure (2.2 bar) and close to the saturation limit (270 bar) exp  130, whereas the ideal gas law gives ideal gas  5.

o

o

270 bar 210 bar 160 bar 80 bar 40 bar 20 bar 10 bar 5 bar

Number of bubbles

400

200

40 C

320 bar 270 bar 210 bar 160 bar 80 bar 40 bar 20 bar 10 bar 5 bar

600

Number of bubbles

22 C

400

200

0

0 2

4

6

8

10

12

14

16

18

20

22

2

24

4

6

8

10

12

14

16

18

20

22

24

Bubble radius, m

Bubble radius, m

Fig. 6. Aphron size histograms for temperature 220C and 400C and various values of final pressure, initial pressure is 450 bar, pressure decrease rate during decompression stage is 800 bar/min.

Temperature effect One of the main quantities that control the nucleation and growth of bubbles in gas supersaturated solution is gas solubility. At fixed amount of gas dissolved in the solution gas solubility determines the part of the dissolved molecules that will eventually migrate to the bubble phase (gas supersaturation). The solubility of gases in liquids is generally characterized by the Henry law which defines how the equilibrium concentration of gas molecules in the liquid phase depends on their concentration in the gas phase.

7

According to the Henry law the equilibrium concentration of the dissolved gas is proportional to the gas pressure at the plane surface of the solution: c = K H·P or, via dimensionless constant, c = s·cg = s·P/R GT, where s = K H·RGT, and RG is the universal gas constant. The Henry constant KH is a temperature-dependent parameter, generally decreasing with temperature. For nitrogen the Henry law gives a good approximation of solubility dependence on pressure up to 100 MPa and 0 temperature up to 250 C [7]. For the known content of the gas entrapped in the aphron foam at ambient conditions (0.84 at 1 0 bar, 22 C) the saturation pressure is defined as the minimal pressure under which all the bubbles get eventually dissolved in the base solution. The saturation pressure for nitrogen in the 0 0 xanthane solution is Ps  320 bar at 22 C and Ps  430 bar at 40 C. The values are estimated from the microscope images in the limit where the size of the stable aphrons approaches the microscope resolution (~1 µm). To determine Ps more accurately one should use e.g. highfrequency acoustic spectroscopy technique which is sensitive to bubbles with size of several tens of nanometers, however high viscosity (~1000 Pa∙s) of the carrying solution may obscure the measurements. The found values of Ps enable us to determine the Henry constants for our base solution and to compare them with constants for pure water. Assuming that the total number of gas molecules is preserved during the gas dissolution, cg·φ = (cs – c)(1 – φ), where φ is the final gas content of the bubble fluid, we can write:

 cs  c    ,  Ps  P   1   

where the gas oversaturation ζ = (Ps – P)/P

is

introduced. From the expression it follows that the product s·ζ is equal to the ratio of the total bubble

, m

s  RG T

10 0

40 C A = 3.8, n = 0.33 0 22 C A = 5.4, n = 0.27

volume to the base fluid volume: Vgb/Vf. By substitution of φ = 0.84 and P = 1 bar we get: s 0

1 0,1

(22 C) = 0.016, and s (40 C) = 0.012. The 0 corresponding values for pure water [7]: s (22 C) = 0

0.0154 and s (40 C) = 0.0125.

1

10

100

(Ps - P)/P

0

Fig.7. Power law fitting for dependence of mean bubble radius on gas super-saturation for 220C and 400C.

Since at higher temperature the solution gets more gas supersaturated the bubble nucleation and the related aphron formation starts at higher pressure. Such behaviour is illustrated by Fig.6 where the variation of the aphron SD with 0

0

pressure is depicted for the temperature values of 22 C and 40 C. It is seen that the aphron SDs are almost the same at the low values of pressure while at the high values they are different. At 0

the maximum pressure value of 320 bar there are no aphrons visible by microscope at 22 C 8

0

while the aphrons are still exist and have the mean radius of 2 m at 40 C. As to the already formed aphrons the effect of heating is twofold: their stability against gas dissolution increases while the stability against drainage decreases due to the reduction in the fluid viscosity. Another important issue is the thermal degradation of the CGA multilayer structure. Despite the fact that the developed microscopy technique is unsuitable to discriminate the layers of the aphron shell at high pressures and an alternative approach based e.g. on X-ray tomography or NMR is required, we still can suppose that the more pronounced deviation of the coarsening 0

behaviour from the Lifschitz-Slyozov-Wagner law (the detailed description follows) at 22 C 0 than at 40 C as seen on Fig.7 is an indication that the properties of the aphron shell are changing with temperature. However further experiments are needed to support the supposition.

Pressure release rate effect In theory the evolution of a supersaturated solution in time is generally subdivided into three stages: nucleation, independent growth of particles of minority phase, and Ostwald ripening (coarsening) where large particles grow at the expense of small ones. Such subdivision is artificial and the stages do overlap yet it gives an insight how the BSD can be adjusted in its parameters. Indeed the BSD is affected the most during the stage of nucleation which defines the initial density and size of gas phase embrions. The nucleation is governed by the gas supersaturation which is a time-dependent quantity as the pressure is usually released at a finite rate. The dependence of the BSD on pressure release rate has been studied in a number of experimental works [8,9], however there is still no analytical description of the effect [11]. The other stages are governed by diffusion and provide scarce opportunities for the BSD control. Assuming the bubble nucleation is homogeneous and neglecting the effect of the vapour pressure the nucleation rate J(t) can be represented in the form [10]: J (t )  Z  exp(   G (t ) kT ) ,

where the free energy ΔG(t) is expressed as: G (t )  4 Rc2 (t ) 3,

Rc (t )  2

 c (t )

K H  P (t )  ,

γ is the surface tension, Rc (t ) is the critical bubble radius, c (t ) is the concentration of the dissolved gas averaged over the fluid volume, for constant pressure release rate: P(t )  Ps  t  dP dt . As it is shown in [11] the function dc (t ) dt decreases monotonically with

time, so that there is point in time where G (t ) has minimum: dc (t ) dt   K H  dP dt . The bubble nucleation occurs in the vicinity of the minimum where J (t )  J th , and the nucleation 9

threshold value is usually accepted as

A = 17, n = 0.13

2E8

1,2E8

3 on the pressure release rate dP/dt

9 0,15

-3

and the released gas content  = (4π/3) NV

10

0,18

8

8E7 7

, m

1,6E8

mean radius , volume concentration NV

NV, cm , 

J th  0.01 J max . The dependence of the

0,12

during decompression stage are represented on Fig.8. For the calculation of the volume concentration NV from the surface concentration NAof an image the authors follow DeHoff [12] and write NV = NA /D, where D is a mean tangent diameter obtained by averaging over all orientations of the projection plane, it is assumed further that D

4E7

6

A = 6.8E6, n = 0.43 100

1000

dP/dt, bar/min

Fig.8. Power law fitting for dependence of bubble density and mean bubble radius on pressure release rate. Conservation of released gas content is demonstrated. Initial pressure is 400 bar, final pressure is 30 bar, temperature is 220C.

= 2. The aphron mean size and the particle concentration dependencies on pressure release rate are well approximated by power law. The scaling laws are consistent: the released gas content  is almost independent of dP/dt.

Coarsening of CGA based fluid In order to study the coarsening of the CGA based fluids the evolution of the aphron SD in time is analyzed for various values of the gas supersaturation during the third stage of the aphron foam preparation. The gas super-saturation ζ = (Ps - P)/ P is a key parameter as it determines the amount of gas that can be released after the gas saturated solution is decompressed to some (a)

5 bar

(b) 0,3

60 bar 90 bar 120 bar 180 bar

10

/

, m

10 bar 30 bar

0,2

5 bar 10 bar 30 bar 60 bar 90 bar 120 bar 180 bar

0,1 2 10

100

1000

10

Time, min

100

1000

Time, min

Fig.9. Dependence of first (a) and second (b) moments of aphron SD on time. Initial pressure is 400 bar, final pressure at decompression end is family parameter, pressure release rate is 1600 bar/min, temperature is 220C. 10

pressure below saturation limit P < Ps. A curious finding the authors encountered in the process of the aphron foam preparation is that the amount of gas (air or nitrogen) entrapped during the high-speed homogenization at ambient conditions (first stage) is sufficient to make the base xanthane solution gas supersaturated at reservoir conditions. This fact along with the aphron little affinity to each other and to mineral surfaces (charged colloid) account for the extraordinary stability of the aphrons at reservoir pressure and temperature despite their micron and submicron size. Fig.9a depicts the evolution in time of the aphron mean radius . Levenberg-

1.5 min 23 min 2 R = 0.86 2 hours 2 R = 0.98

Normalized bubble density

1,0

Marquardt curve fitting performed with power law model (t) = A·tn gives for the exponent n values close to 0.28 (for all of the curves within 3% scatter). Common theories of the bubble growth in gas supersaturated solution [13,14] give for the exponent n = 1/2

0,5

0,0 4

8

12

Bubble radius, m

as they are based on the assumption that the dissolved gas concentration is constant and the gas diffuses only from the solution to the

Fig.10. Time evolution of aphron SD. Fitting is performed with LSW asymptotic function. Initial pressure is 400 bar, final pressure is 90 bar,

bubbles and thus they correspond to the second stage of the BSD evolution. However the appreciable growth of the bubble critical radius with the depletion of the dissolved gas results in the reverse process: the replenishment of the solution with the gas caused by the dissolution of the large-size subcritical bubbles. Such theory was first proposed by Lifschitz and Slyozov [15] and independently by Wagner [16] to characterize diffusive decomposition of solid solutions and later was it noticed that similar equations describe the coarsening of wet foams [14]. For the mean/critical radius growth the LSW theory gives for the exponent n = 1/3. Recent theoretical studies show that despite the fact that the LSW theory corresponds to the limiting case of vanishing gas volume the exponent value is preserved for the entire range of the gas content (equilibrium value as t

is

assumed), however the asymptotic SD may differ considerably from the LSW one [17-19]. The value of 0.28 obtained for the aphron SD coarsening is very close to that of the LSW theory. Moreover the authors fitted the experimental aphron BSD with the asymptotic function of the LSW theory and got fine matching for SDs older than 2 hours though the equilibrium gas content is 0.84 (see Fig. 10). The dissolution of large bubbles can be tracked on the Fig.11, where the instantaneous radius of individual bubbles is plotted versus time (size trajectories). It is seen that even those bubbles that were lucky to grow at the expense of their neighbours that vanished earlier eventually dissolve as well despite their large size. 11

Fig.9b depicts the evolution in time of the normalized standard deviation / which represents the dispersity of the aphron colloid. A remarkable fact of the CGA based fluid is that

60 50 40

the dispersityis an almost conserved quantity: it Rk, m

is independent of pressure and it changes only by two times before the wet foam reaches the

30

scaling state (where it is independent of time). There is a tendency that for the smaller values of

20

10

the gas supersaturation (higher final pressure) the better monodispersity is observed, however it is probably the consequence of the shorter period of decompression (as the pressure release rate is

10

100

1000

Time, min

Fig.11. Instantaneous radius of individual bubbles versus time.

constant).

Conclusions The microbubble based fluids produced by the developed method have narrow SD and are capable to withstand pressure and temperature conditions close to those of a reservoir. The coarsening of such wet foams is described by the LSW law. After several minutes from the preparation the dispersity of the colloid remains constant in time. The question whether the microbubbles preserve the multilayer structure of the gas aphron under such severe conditions is still open as by applying the high-pressure microscope technique developed it is impossible to discriminate the layers of the bubble shell and methods with higher resolution are required. A valuable finding is that the amount of gas entrapped into the base solution during the mechanical homogenization at ambient conditions (foam preparation stage) is sufficient to obtain the gas-supersaturated base solution at reservoir conditions. The microbubble fluids produced from the CGA foam at reservoir conditions can be successfully used as pressure- and temperature- controllable systems for various oilfield applications. REFERENCES 1) Contrast Media in Ultrasonography. Basic Principles and Clinical Applications. – Ed: E. Quaia, Springer-Verlag Berlin, Heidelberg, 2005, 401 p. 2) G. Chen, D. Burnett. Improving Performance of Low Density Drill in Fluids with Hollow Glass Spheres, - SPE 82276, 2003, pp. 1-10. 3) F.B. Growcock, A. Belkin, M. Fosdick, M. Irving, B. O’Connor, T. Brookey. Recent Advances in Aphron Drilling-Fluid Technology. – SPE 97982, 2007, pp. 74-80. 4) P. Jauregi, G.R. Mitchell, J. Varley. Colloidal Gas Aphrons (CGA): Dispersion and Structural Features. - AIChE J., 2000, vol.46, pp. 24-36. 12

5) T. Lappalainen, J. Lehmonen. Determinations of bubble size distribution of foam-fibre mixture using circular hough transform. - Nordic Pulp and Paper Research J., 2012, vol.27, pp. 930-939. 6) G. Bradski, A.Kaehler. Learning OpenCV: Computer vision with the OpenCV library. O’Reilly Media Inc., 2008, p.158. 7) A. Yu. Namiot. Solubility of gases in water, 1991, Nedra, Moscow, 86 p. (in Russian) 8) Q. Guo, J. Wang, Ch.B. Park, M. Ohshima. A microcellular foaming simulation system with a high pressure-drop rate. – Ind. Eng. Chem. Res., 2006, vol.45, pp.6153-6161. 9) S.N. Leung, A. Wong, Ch.B. Park, Q. Guo. Strategies to estimate the pressure drop threshold of nucleation for polystyrene foam with carbon dioxide. – Ind. Eng. Chem. Res., 2009, vol.48, pp.1921-1927. 10) J. Frenkel. Kinetic Theory of Liquids. - Dover, New York, 1955 . 11) K. Taki. Experimental and numerical studies on the effects of pressure release rate on number density of bubbles and the bubble growth in a polymeric foaming process. – Chem. Eng. Sci., 2008, vol.63, pp.3643-3653. 12) R.T. DeHoff, F.N.Rhines. Determination of the number of particles per unit volume from measurements made on random plane sections: the general cylinder and the ellipsoid. Trans AIME, 1961, vol.221, pp.975-982. 13) A.E. Kuchma, F.M. Kuni, A.K. Shchekin. Nucleation stage with nonsteady growth of supercritical gas bubbles in a strongly supersaturated liquid solution and the effect of excluded volume. – Phys. Rev. E, 2009, vol.80, pp.1-7. 14) P. Stevenson. Inter-bubble gas diffusion in liquid foam. – Current Opinion in Colloid & Interface Science, 2010, vol.15, pp.374-381. 15) I.M. Lifschitz, V.V.Slyozov. The kinetics of precipitation from supersaturated solid solutions. – J. Phys. Chem. Solids, 1961, vol.19, pp.35-50. 16) C. Wagner. Theorie der Alterung von Niederschlagen durch Umlösen (Ostwald-Reifung). Z.ElektroChem, vol.65, 1961, pp. 581-591. 17) P.V. Voorhees. The theory of Ostwald ripening. – J. Stat. Phys., 1985, vol.38, pp.231-252. 18) P. Streitenberger. Analytical description of phase coarsening at high volume fractions. – Acta Mater., 2013, vol.61, pp.5026-5035. 19) J. Svoboda, F.D. Fisher. Generalization of the Lifshitz–Slyozov–Wagner coarsening theory to non-dilute multi-component systems. – Acta Mater., 2014, vol.79, pp.304-314.

13