J. Phys. Chem. B 2008, 112, 10455–10465
10455
Alcohol Cosurfactants in Hydrate Antiagglomeration J. Dalton York and Abbas Firoozabadi*,† Department of Chemical Engineering, Mason Laboratory, Yale UniVersity, New HaVen, Connecticut 06520 ReceiVed: February 27, 2008; ReVised Manuscript ReceiVed: June 20, 2008
Because of availability, as well as economical and environmental considerations, natural gas is projected to be the premium fuel of the 21st century. Natural gas production involves risk of the shut down of onshore and offshore operations because of blockage from hydrates formed from coproduced water and hydrateforming species in natural gas. Industry practice has been usage of thermodynamic inhibitors such as alcohols often in significant amounts, which have undesirable environmental and safety impacts. Thermodynamic inhibitors affect bulk-phase properties and inhibit hydrate formation. An alternative is changing surface properties through usage of polymers and surfactants, effective at 0.5 to 3 weight % of coproduced water. One group of low dosage hydrate inhibitors (LDHI) are kinetic inhibitors, which affect nucleation rate and growth. A second group of LDHI are antiagglomerants, which prevent agglomeration of small hydrate crystallites. Despite great potential, work on hydrate antiagglomeration is very limited. This work centers on the effect of small amounts of alcohol cosurfactant in mixtures of two vastly different antiagglomerants. We use a model oil, water, and tetrahydrofuran as a hydrate-forming species. Results show that alcohol cosurfactants may help with antiagglomeration when traditional antiagglomerants alone are ineffective. Specifically, as low as 0.5 wt. % methanol cosurfactant used in this study is shown to be effective in antiagglomeration. Without the cosurfactant there will be agglomeration independent of the AA concentration. To our knowledge, this is the first report of alcohol cosurfactants in hydrate antiagglomerants. It is also shown that a rhamnolipid biosurfactant is effective down to only 0.5 wt. % in such mixtures, yet a quaternary ammonium chloride salt, i. e., quat, results in hydrate slurries down to 0.01 wt. %. However, biochemical surfactants are less toxic and biodegradable, and thus their use may prove beneficial even if at concentrations higher than chemical surfactants. 1. Introduction Because of supply and clean-burning features, natural gas may occupy a larger share of energy consumption and surpass oil in the 21st century. There is less CO2 produced from burning natural gas. Furthermore, produced CO2 can be separated from the combustion products and sequestered more readily than oil combustion products for improved CO2 recovery or storage in saline aquifers. However, conditions existing in gas production lines often favor formation of crystalline-inclusion compounds known as gas hydrates. Water, coproduced with natural gas, forms lattice structures by hydrogen bonding; the structures are stabilized by guest molecules such as methane, propane, etc., under high pressures and temperatures in the range of a few degrees to 25 °C.1 Formation of gas hydrates occurs rapidly, unlike corrosion, scale, or wax buildup. This rapidity has undesirable safety and environmental consequences. Plug formation from hydrates may lead to production shutdowns. Traditional hydrate prevention methods include physical means, such as insulation and electrical heating. Thermodynamic inhibition through change of bulk-phase properties with methanol (MeOH) and monoethylene glycol (MEG) is widely used. Thermodynamic inhibitors shift equilibrium conditions to lower temperature and higher pressure.2 Although well-characterized, these inhibitors often require large concentrations, as high as 60 weight (i.e., wt.) % of coproduced water, which increases costs and has serious environmental impact.3 * Corresponding author. E-mail:
[email protected]. † Affiliated with the Reservoir Engineering Research Institute, Palo Alto, CA 94306.
An alternative to thermodynamic inhibitors is the use of lowdosage hydrate inhibitors (LDHI). LDHI mainly influence hydrate surface properties and are effective at concentrations of 0.5 to 3 wt. %; rather than affecting thermodynamic equilibrium, they act upon kinetics or agglomeration. Kinetic hydrate inhibitors (KHI) are generally polymeric compounds that delay nucleation and decrease growth rate.4,5 KHI may result in complete inhibition of hydrates6 but do not perform well at pipeline/well shut-in conditions or at high operating subcoolings, i.e., ∆Top, the difference between equilibrium temperature and operating temperature at a given pressure. It should be noted that ∆Top is not the same as the difference between the equilibrium (dissociation) temperature and the crystallization temperature. In this work we refer to this second form of subcooling as the onset subcooling, denoted by ∆Ton. In some flow conditions, ∆Top may be as high as 20 °C; therefore, one would require effective LDHI for ∆Top ) 20 °C. Shut-in conditions, i.e., when pipeline flow is paused for a period of time, may occur when pipeline/well maintenance is necessary or when inclement weather occurs. A second class of LDHI are antiagglomerants (AA) which prevent agglomeration, but not formation, of hydrate crystals and enable hydrate transportation as slurries. AA are generally effective at high ∆Top or at shut-in conditions.7–9 AA may also possess kinetic inhibition features.10–12 They are generally surfactants but may be low molecular weight oligomeric species.9,13 AA have not been studied as extensively as KHI. Insight into hydrate antiagglomeration and mechanism are found in surfactant and colloidal science literature.14–16 AA structure is key to their effectiveness and mechanism.17 Effective AA contain the headgroup that can interact with a
10.1021/jp8017265 CCC: $40.75 2008 American Chemical Society Published on Web 07/31/2008
10456 J. Phys. Chem. B, Vol. 112, No. 34, 2008 water lattice, such as amine or carbonyl groups, through hydrogen bonding or electrostatic attraction. AA compounds may also contain head groups that act as hydrate guest molecules, this feature combined with hydrogen bonding may incorporate the AA into crystals. Molecules in this case may adsorb too strongly and become engulfed in the growing crystal, requiring higher concentrations. The hydrophobic tail renders hydrate more oil-wet, thus dispersible in the oil phase, and prevents separate crystals from agglomerating.18 AA often produce water-in-oil (w/o) emulsions, thus limiting hydrate growth to water droplets dispersed in the oil phase.9,17,19–24 However, emulsion stability is generally undesired in gas and oil production.25,26 Once transportation of well fluids is complete, it is desired that phase separation be attainable so that product quality standards can be met. If these emulsions are too stable, then additional processing or additives may be required once hydrate formation is of no concern. AA may become ineffective if water occupies one-third or more of the combined volume of water and oil in the process stream. That is, their effectiveness may be limited to mixtures of 50% water-cut, defined as the ratio of water volume to that of oil, or less. This requirement may be related to w/o emulsion formation, but other reasons such as high slurry viscosity with high hydrate volume fraction is also cited in the literature.17,19,27 One recent study28 showed that a newly developed AA inhibitor is effective at water cuts as high as 80% in both wet-tree and dry-tree applications, but this is currently the only known exception to the 50% water-cut limitation. In most gas production flow-lines, the amount of hydrocarbon liquid is more than the amount of coproduced water, and therefore the generation of w/o emulsions may not be an issue. However, in some cases, water production may be high, and therefore the study of varying fluid composition on antiagglomerant performance becomes of interest. Most antiagglomeration studies9,20,21,29 use a relatively low, and constant, water-cut. One study reported variable watercuts, as high as 80%; however, the crude oil used also contained the hydrate-forming components so there was less hydrate formed at high water-cuts.30 Alcohol cosurfactants are widely discussed in the literature. It is well-known that cosurfactants aid in microemulsion formation, by interacting with primary surfactant in the interfacial region in reducing oil/water interfacial tension.31 There is also evidence of cosurfactant effects such as modifications in primary surfactant packing and head area, reduction in interfacial layer thickness, and variation in continuous-phase viscosity.32,33 The effect of different alcohols has been studied, and it is found that medium-chain alkanols may be the most efficient cosurfactants, yet the smaller chain species, such as MeOH, are also effective.31,34 There are limited data showing that an alcohol affects the performance of kinetic inhibitors. Two studies show contrasting effects of MeOH in mixtures with KHI: synergy can be concluded in one study13 while the other35 shows a well-studied inhibitor, PVCap, is less effective in MeOH’s presence. One theoretical study has examined the effect of MeOH as a cosolvent with very little effect on antiagglomeration.36 To the best of our knowledge, there has been no experimental examination of the effect of alcohol cosurfactants in hydrate antiagglomeration. Very recently there has been interest in biosurfacants as antiagglomerants for ice37 and hydrates.38,39 Biosurfactants are often superior to chemical surfactants because of (1) higher biodegradability, (2) lower toxicity, and (3) safety.40,41 To the best of our knowledge, there are only two reports of the use of
York and Firoozabadi
Figure 1. Multiple screening-tube rocking apparatus.
biosurfactants in the hydrate literature.38,39 In ref 38, 500 ppm of a rhamnolipid surfactant exhibited AA ability but only for one of the oils used in the study. Because of the presence of natural surfactants in the oil, results from ref 38 may not apply to hydrates in natural gas systems. In ref 39, we showed that rhamnolipid at concentrations down to 0.05 wt. % is effective as AA in tetrahydrofuran (THF) mixtures with sufficient model oil, i.e., four parts by weight or higher. Chemically, the major rhamnolipids are glycosides of rhamnose (6-deoxymannose) and B-hydroxydecanoic acid. Rhamnolipids are known to reduce surface and interfacial tension42 and have been used to create stable microemulsions.43 Typical commercial products consist of both the monorhamnolipid and dirhamnolipid forms and are generally more expensive than the chemical counterparts. This work utilizes a multiple screening-tube rocking apparatus to investigate (1) the effect of increasing the water-cut, and (2) the use of alcohol cosurfactants on hydrate antiagglomeration. The influence of AA concentration, ∆Top, and residence time at ∆Top are other variables of focus. Rhamnolipid biosurfactant and a quaternary ammonium salt (quat) are the AA, the same surfactants used in our previous work.39 Shut-in testing in which vials with hydrates are allowed to stand unagitated for a given period, and emulsion stability tests are also included. Through these variables, model w/o emulsions are judged according to hydrate formation/dissociation temperatures and visual observations of agglomeration state after hydrate formation. THF is used as the guest molecule, since it forms structure II hydrates at atmospheric pressure, the same type that forms in most pipelines.44 There are several differences between THF and real systems, but THF is still considered to be an adequate model system. THF may partition significantly between the aqueous and organic phases.21 Another major difference is higher solubility of THF in water than any species found in a typical natural gas mixture. THF and some gases, e.g., CO2, may initiate hydrate in the bulk water phase;45–47 however, some authors present data and show methane, methane-ethane, and methane-propane hydrates form at the water/oil interface.47,48 Effective antiagglomerants will reside at or near the interface, where growth occurs in a real hydrate system. In any system, AA shown effective for THF systems can also be effective for systems where hydrate formation and growth occur at the interface. Furthermore, the difference between hydrate growth in THF and real systems is minimized in emulsified mixtures.21 2. Experimental Methods A. Apparatus. The experimental setup, a multiple screeningtube rocking apparatus, is shown in Figure 1, the same as in
Alcohol Cosurfactants in Hydrate Antiagglomeration our previous work.39 It consists of a motor-driven agitator, with a rack holding up to 20 separate borosilicate glass scintillation vials with dimensions of 17 (diameter) by 60 (height) mm, submerged in a temperature bath. Each vial holds roughly 7.4 mL of a test mixture and a ∼8 mm diameter stainless steel 316 ball to aid agitation as well as visual observations; a Teflonlined plastic screw-cap is used along with Teflon tape, around threads, to seal vials. The rack rotates the vials 150° to either side of the vertical direction, completing a cycle every 5 s. The temperature bath used is a Huber CC2-515 vpc filled with 10 cSt, at 24 °C, silicon oil from Clearco Products Co., Inc., Bensalem, PA. Thermocouples, with accuracy of ( 0.2 °C from 70 °C down to -20 °C, are attached to the outside of the vials when crystallization and melting data are desired. These thermocouples must be attached to the outside with the use of such vials, which are ideal for sample preparation and containment. Our data, and comparisons to previous work in the same mixture types,21,39 shows this method is accurate. A different approach to sample preparation and containment, one that involved the insertion of thermocouples inside the sample mixtures, was attempted initially, but there were problem with repeated testings. We suspected that thermocouples served as nucleation active sites. Various testing showed that the inside temperature and outside wall temperature was the same, and then we opted for outside wall temperature in our work. An Agilent 34970A data acquisition unit, recording temperature every 20 s, and ice bath as fixed junction reference temperature is used with all thermocouples. Additional equipment used is as follows. Agglomeration state images are obtained with a ∼169 mm rigid borescope, Hawkeye Pro Hardy from Gradient Lens Corp., Rochester, NY, and Nikon Coolpix 5400 digital camera with samples still in bath fluid. High-accuracy density measurements, to examine THF partitioning, were obtained with an Anton Paar DM 4500, having an accuracy of 1 × 10-6 g/cm3. B. Chemicals. In all test mixtures, deionized water, obtained from a Barnstead Nanopure Infinity system with quality of roughly 5.5 × 10-2 µs/cm, and 99.5%+ purity THF (from Acros) are used. The oil phase consists of 99% purity 2,2,4trimethylpentane (i.e., isooctane, from Acros). The following surfactants are used: rhamnolipid (product JBR 425) was obtained from Jeneil Biosurfactant Co., Madison, WI. ARQUAD 2C-75, dicetyldimethylammonium chloride, was obtained from Akzo-Nobel. It consists of 75 wt. % surfactant in solvent consisting of water (at 5-10 wt. %) and 2-propanol (at 15-20 wt. %). Both were used as supplied. All the above chemicals used are the same as discussed in our previous work.39 As cosurfactant, 99.8% anhydrous MeOH with less than 0.05 ppm water was obtained from Acros. C. Procedure. The experimental procedures are the same as in our previous work.39 A composition of mostly 1/1/2/x parts, x is for varying surfactant concentration in different tests, by weight of water/THF/isooctane/surfactant is employed. Stoichiometric ratio of 0.235/1 for THF/water may lead to ice formation from heterogeneous nucleation.49 Higher concentration of THF avoids ice formation.50 Surfactant concentrations of 1.5 wt. % and less were used, since it was observed previously39 that the desired outcome, i.e., hydrate slurries at large subcooling, was achieved at such low concentrations; it is thus thought that no benefit would come from using higher surfactant concentrations. THF partitioning, between the water and oil phases, in a mixture of two parts isooctane was examined through highaccuracy density measurements. Such a test was executed by
J. Phys. Chem. B, Vol. 112, No. 34, 2008 10457
Figure 2. Typical freeze-thaw cycle data for a mixture of two parts isooctane and no MeOH; example for a mixture of 1.5 wt. % rhamnolipid.
mixing a sample of only 1/1/2 parts by weight of water/THF/ isooctane and allowing for equilibration of concentration for a period of 3 days at roughly 25 °C, i.e., room temperature in our laboratory. The density of the aqueous portion of this mixture was then measured at 25 °C. Several known concentrations of THF in water were also measured to supply a calibration for the test mixture. More extensive data for THF partitioning in a mixture of four parts isooctane are available in the literature.21 In our measurements a single temperature was employed for comparison. When MeOH as a cosurfactant is effective in preventing agglomeration, a systematic series of tests are conducted to examine the limits of both MeOH and AA concentration required in antiagglomeration. MeOH concentrations of 5, 2, 0.5, and 0.1 wt. % are employed in our study. Limited agglomeration state testing was conducted with zero or one part isooctane; in these cases, up to 10 wt. % MeOH and only 1.5 wt. % AA is employed to examine the effect on agglomeration. MeOH in the amount of 10 wt.% is not used extensively in this study because much lower concentrations prove effective. Also, the effect of methanol cosurfactant on mixtures of four parts isooctane and very low rhamnolipid concentration, concentrations that resulted in agglomeration in our previous work,39 were tested for agglomeration state only. The water-cut values in all mixtures in our work refer to the ratio of the water volume to that of isooctane, i.e., these values do not include amount of THF or MeOH. Temperature data was acquired separate from visual observations, since half of the sample vial surface area is covered when thermocouples are attached. However, temperature control of the bath and agitation are the same for both types of experiments. Kinetic/Thermodynamic Data Acquisition. Select mixtures with 1/1/2/x of water/THF/isooctane/surfactant parts by weight (some mixtures also include 5, 2, or 0.5 wt. % methanol) were tested in the following manner. A typical trend of measured temperature data, referred to as a freeze-thaw cycle, is shown in Figure 2. Mixtures are brought to 7 °C, allowed to reach equilibrium, and then a 10 °C/h cooling ramp is employed to -8 °C. The temperature is then raised back to 7 at 14 °C/h. A temperature of 7 °C is chosen because it is used in a previous study.21 The hydrate equilibrium temperature for these mixtures with substantial THF partitioning in the oil is around 4 °C as shown below. As a mixture is cooled below the equilibrium- or dissociationtemperature, an onset of hydrate crystallization occurs and an exothermic heat release begins. The temperature of this transition
10458 J. Phys. Chem. B, Vol. 112, No. 34, 2008
Figure 3. Typical freeze-thaw cycle data for mixtures with two parts isooctane and 5 or 2 wt. % MeOH; example for a mixture with 0.5% rhamnolipid and 5% MeOH.
Figure 4. Typical freeze-thaw cycle data for mixtures with two parts isooctane and 0.5 wt. % MeOH; example for a mixture with 0.05% quat.
is the crystallization temperature, or Tc. After crystallization, the sample temperature rejoins that of the bath fluid. The time, tc, that the mixture spends crystallizing is directly related to the growth rate. Dissociation of the mixture upon heating shows as an endotherm, the beginning of which is labeled the dissociation temperature, Td. The accuracies of individual Tc and Td data are determined at ( 0.2 and ( 0.5 °C, respectively. The accuracy of tc data is (1.0 min. With 5 or 2 wt. % MeOH in solution, tc and Td data become less clear than in all other instances, as shown in Figure 3. In this case, tc can still be discerned within (1.0 min accuracy even though the crystallization peak does not end as abruptly. However, the accuracy in determining Td suffers after addition of MeOH, again, for 5 and 2 wt. %. In these cases, the accuracy is more like (0.7 °C, although the determination is easier when the plot is enlarged. When only 0.5% MeOH is added, the effect on accuracy of tc and Td is less than for higher MeOH amounts; yet the effect is slightly more than when no MeOH is present, shown in Figure 4. In these cases, the crystallization peak ends more gradually than is shown in Figure 2, but tc may still be determined within (1.0 min accuracy. Each composition was prepared in triplicate and experiments were repeated five times per sample. Thus, each sample was reused for five consecutive experiments. Some tests were separated by periods of heating at 7 °C for 20 min. In other cases, unagitated samples were kept in the bath overnight as it gradually warmed to room temperature before proceeding to the next test. There is no difference between the results from the use of samples exposed to room temperature and to those limited to heating at 7 °C. Data shown below is the average of 15 separate experiments per composition.
York and Firoozabadi Agglomeration State. Experiments for visual observations were performed similarly to crystallization/dissociation testing. Agitated mixtures are equilibrated at 7 °C and then 10 °C/h cooling is applied to bring the mixtures to a minimum temperature of -8, -12, -16, or -20 °C to examine the effect of large operating subcooling, ∆Top. The procedure deviates here whereby the minimum temperature is held constant for 24 h. Observations are made at 10 min, 1 h, and 24 h into this period. This was repeated twice for each of the triplicate mixtures for a given composition, separated by periods of heating at 25-30 °C for 30 min. There is no dependence of our results on this heating temperature. Observations were made mainly with the naked eye but also with the borescope. These observations show whether a dispersion, i.e., slurry, of hydrate crystals created by the surfactant or agglomeration occurs. Shut-in testing was also conducted. All mixtures were left unagitated at the various ∆Top for 60 min. This test was conducted at the termination of the cooling ramp when the desired temperature was reached. Agitation was then resumed to assess whether hydrates are redispersed or irreversible agglomeration occurred. Emulsion Stability. The procedure for emulsion stability testing is similar to that employed by Zanota et al.21 A similar approach is used in other hydrate antiagglomeration studies.30 Select mixtures with 1/1/2/x of water/THF/isooctane/surfactant parts by weight (some mixtures also include 5, 2, or 0.5 wt. % methanol) were prepared without the stainless steel ball and homogenized by hand shaking for 1 min; the rate and motion of this agitation was the same as would be imparted by the multiple screening-tube rocking apparatus. The fluid was transferred at room temperature to a graduated cylinder, with glass stopper, and the time for separation of 60 volume % of the initial aqueous phase was measured and used as indicator of emulsion stability. This procedure is referred to as “fresh” sample emulsion stability testing. Tests at each composition were repeated three times. Alternative assessment of emulsion stability, referred to as “used” sample emulsion stability testing, was performed in the following manner. After a visual observation test was finished, mixtures were allowed to equilibrate at room temperature. Samples were removed from bath and agitator, after which handshaking was applied for 1 min, using the same rate and motion as used with “fresh” samples. The mixtures were then monitored for phase separation, with an approximate indicator of the 60 volume % separation being marks made on the vial during mixing that also corresponds to the height of the steel ball. Tests at each composition were repeated three times. Note that one difference between the two types of tests is the presence of the steel ball in the latter. Also, 1 min of agitation was enough time to fully emulsify in both types of tests. Both types of tests were performed at room temperature for the following reasons. These results are primarily relevant to potential emulsion stability of downstream fluids, presumably close to ambient temperature, in an actual production process. The claim has also been made21 that surfactants used to generate more stable emulsions in these model mixtures at room temperature are more effective as antiagglomerants. Thus, emulsion stability under other types of conditions was not considered. 3. Results A. Agglomeration State. Preliminary agglomeration state testing began with a few mixtures of 1/1/4/x parts by weight of water/THF/isooctane/rhamnolipid, as continuation of previous
Alcohol Cosurfactants in Hydrate Antiagglomeration
Figure 5. Agglomeration state results for mixtures with four parts isooctane and small amounts of rhamnolipid. In all cases, significant adhesion of hydrate upon vial walls occurs immediately at all minimum temperatures (represented by • symbol). Data represents behavior of a given composition across all minimum temperatures.
Figure 6. Significant adhesion observed in mixtures of four parts isooctane, very low concentrations of rhamnolipid, and 5 wt. % MeOH or less; data shown in Figure 5. Sample shown in image contains 0.01 wt. % rhamnolipid and 5 wt. % MeOH.
work.39 As shown in that paper, these mixtures begin to show significant hydrate adhesion upon vial walls at rhamnolipid concentrations of 0.01 wt. % and below. In this work, it was thought that small amounts of MeOH cosurfactant might allow full hydrate slurries to persist at lower rhamnolipid concentrations. As shown in Figure 5, up to 5 wt. % of MeOH does not alleviate the adhesion problem. An example of such adhesion is shown in Figure 6. It appears MeOH does reduce the magnitude of adhesion, but it is still considered significant. Data shown in Figure 5, and in all of the agglomeration state plots in this paper, represent the behavior of a given composition across the entire temperature range tested. In all cases, this behavior is consistent across all minimum temperatures, i.e., at -8, -12, -16, and -20 °C. Some differences do occur for rhamnolipid concentrations, in data for both four and two parts isooctane mixtures, exhibiting adhesion, where there is typically increased adhesion for lower temperatures tested. Also, mixtures of four parts isooctane without AA, but with 0.5 and 5 wt. % MeOH, respectively, were tested across the same temperature range. The results were plugs in all tests. Since the point of this study was to examine effects of MeOH combined with AA, and not the effect of MeOH alone, in hydrate mixtures, these data are not included in the plots shown in this paper. Also,
J. Phys. Chem. B, Vol. 112, No. 34, 2008 10459
Figure 7. Typical plug appearance when small amounts, i.e., zero to two parts by weight, of isooctane are used in mixture. In this image, the vial is upside down with most of the vial volume blocked by hydrate; the mixture being tested in this image is one containing two parts isooctane. Calculations using measured THF hydrate density58 show the volume fraction of hydrate in mixtures of two parts isooctane is roughly 0.25. This compares to the calculated volume fraction of 0.16 in mixtures of four parts isooctane.39 The sample shown in this image is for a mixture of two parts isooctane and 1.5 wt. % quat. Small bubbles seen in this image, as well as in Figure 15, are present in the bath fluid due to bath operation.
similar tests were not conducted on mixtures with two parts isooctane since the results would likely be the same. Next we studied the effect of increasing water-cut upon antiagglomeration. Mixtures exhibiting effective antiagglomeration with four parts isooctane were used as a reference. Mixtures with zero isooctane and 1.5 wt. % of either AA formed plugs immediately at -8 °C and below. The same mixtures, but with 5 and 10 wt. %, respectively, MeOH added, also result in plugging behavior at -8 °C and below. The same outcome is found in mixtures of one part isooctane with 1.5 wt. % AA and 0, 5, or 10 wt. % MeOH. An example of plugs in these cases, where the plug occupies most or all of the vial volume, is shown in Figure 7. The 10 wt. % MeOH is not used extensively in this study but only as a probe to determine if higher concentrations of MeOH, i.e., beyond 5 wt. %, would affect antiagglomeration. Thus it is determined that higher concentrations may not provide additional benefit. It was not until mixtures of two parts isooctane were tested that the effectiveness of MeOH cosurfactant was revealed by our method. Figure 8 shows the results of such tests with rhamnolipid as AA. Plugs, either full or partial where the steel ball is blocked from moving across the entire length of the vial, are still observed when up to 1.5 wt. % rhamnolipid is added without MeOH. However, when just 0.5 wt. % MeOH is added to these mixtures, hydrate slurries are formed. The same is seen with up to 5 wt. % MeOH. An example of such slurries is shown in Figure 9. A significant difference exists over the slurries seen in our previous work39 due to increased hydrate volume in the current work. However, there is some agglomeration when a very low concentration of 0.1 wt. % MeOH cosurfactant is used (see Figure 8). Figure 10 shows the results in mixtures with the quat where the AA alone does not produce slurries, but as little as 0.5 wt. % MeOH cosurfactant does. However, a 0.1 wt. % MeOH concentration does not result in antiagglomeration. All mixtures exhibiting stable hydrate slurries passed shutin testing. These results are for all minimum temperatures. That is, mixtures that initially contain hydrate slurries at any minimum temperature are able to have the hydrate redispersed after agitation is paused for 60 min and then resumed.
10460 J. Phys. Chem. B, Vol. 112, No. 34, 2008
York and Firoozabadi
Figure 8. Agglomeration state results for mixtures with two parts isooctane and rhamnolipid with and without MeOH: (9) stable dispersion, i.e., effective antiagglomeration, (•) immediate and significant adhesion upon vial walls, (×) plugging tendency. Plugging tendency means that either total, i.e., steel ball is unable to move, or partial, i.e., steel ball is unable to move through entire length of vial, plugs occur. Data represents behavior of a given composition across all minimum temperatures.
Figure 9. Image of hydrate slurry in mixtures with two parts isooctane. This image was taken with the vial almost horizontal in agitator rack and thus prior to complete slurry settling; because of high hydrate volume present in vial, it was not possible to capture a clear image of the slurry separate from the oil phase. This mixture shown here is for 0.5 wt. % rhamnolipid and 2 wt. % MeOH cosurfactant.
B. Kinetic/Thermodynamic Characteristics. Figures 11 to 14 show results of freeze-thaw cycles for select mixtures of both rhamnolipid and quat with 5, 2, and 0.5 wt. % MeOH, as well as mixtures without AA and/or MeOH. For both surfactants, AA was used in the amounts of 1.5, 0.5, and 0.05 wt. %. The difference between the dissociation temperature Td and the crystallization temperature Tc is the onset subcooling denoted by ∆Ton. C. Emulsion Stability. Tables 1 and 2 provide emulsion stability results for quat and rhamnolipid, respectively, along with standard deviations. Mixtures of select AA concentration with 5, 2, 0.5, and 0 wt. % MeOH were tested using both “fresh” and “used” samples. For both surfactants, AA was used in the amounts of 1.5, 0.5, and 0.05 wt. %. Average and standard deviations are given to the nearest 0.1 min because of the relative instability of most compositions tested. 4. Discussion A. Agglomeration State. Regarding the effectiveness of MeOH as a cosurfactant, low concentration is highly desirable. This is analogous to the LDHI concept, in that inhibitors effective at low concentrations should be used to reduce costs and other impact. Thermodynamic inhibitors, especially MeOH,
may give rise to salt precipitation in petroleum fluid mixtures,,51 and so it is crucial in this respect to be able to identify low concentrations at which MeOH cosurfactant will be effective. When mixtures of zero part isooctane were tested, one sample with 1.5 wt. % rhamnolipid and 10 wt. % MeOH exhibited slurry consistency for a few minutes at -8 °C. In these samples, visual inspection was performed immediately at the minimum temperature in addition to other times mentioned in the procedure. It soon appeared that the steel ball was partially blocked in the vial, although the appearance of the dispersion remained the same, i.e., it did not look like a plug but rather still appeared as a highly concentrated hydrate slurry (see Figure 15). It is thought that the initial blockage in this case is due to the shear volume of hydrate packed to the extent that the steel ball cannot pass. All other samples of this composition, even for 1.5 wt. % quat, never exhibited slurry behavior at any temperature, but the blockages appeared the same; occasionally, the steel ball was unable to move so full plugging does occur. The highly viscous nature of the hydrate slurry in this case appears to play a major role in such blockages. These observations have been made elsewhere.27 Surprisingly, mixtures of one part isooctane always appear clearly as a full hydrate plug, even when 10 wt. % MeOH is added; the same can be said for mixtures of two parts isooctane with surfactant and no MeOH. The results showing plugging tendency in mixtures of zero or one part isooctane are expected. If no emulsion is created in the mixture, i.e., in the case with no oil, then hydrate antiagglomeration may not be sustained. Additionally, hydrate slurries may not be sustainable if insufficient oil is present, perhaps owing to oil-in-water emulsion formation or simply the inability for all hydrate formed to be dispersed within the oil phase. Work has been reported on antiagglomeration in pure water with NaCl as well as low subcoolings, i.e., a few degrees or less. Under suchconditions,dispersionofcrystallitesinwaterareobserved.37,52,53 These limitations are exacted in such studies to control crystal growth, the point of these studies being to create transportable ice slurries for latent heat storage. In a hydrate system, this is not an option, simply due to the impossibility of controlling conditions in the vicinity of a pipeline or well, and so care must be taken to limit hydrate slurry volume and viscosity. Water-cuts much higher than 50% may cause agglomeration in the model mixtures in our work. Mixtures with one part isooctane contain a water-cut of roughly 69%. We expect blockages to form for such as system. However, mixtures of
Alcohol Cosurfactants in Hydrate Antiagglomeration
J. Phys. Chem. B, Vol. 112, No. 34, 2008 10461
Figure 10. Agglomeration state results for mixtures with two parts isooctane and ARQUAD 2C-75 with and without MeOH: (9) stable dispersion, i.e., effective antiagglomeration, (•) plugging tendency. Plugging tendency means that either total, i.e., steel ball is unable to move, or partial, i.e., steel ball is unable to move through entire length of vial, plugs occur. Data represents behavior of a given composition across all minimum temperatures.
Figure 11. Average of freeze-thaw cycle data for mixtures of two parts isooctane with or without rhamnolipid. Given are Tc (shown as ν), tc (shown as columns), and Td (shown as 9). Error bars are presented for all points; some may not be clear because of magnitude.
two parts isooctane contain roughly 34% water-cut, and results still show plug formation when primary surfactant alone is present. One benefit of MeOH cosurfactant manifests itself clearly in these experiments, i.e., when water-cut is increased and agglomeration state is assessed visually. MeOH cosurfactant, and other alcohols, may reduce emulsion droplet and hydrate crystallite size. In our work, small amounts of MeOH cosurfactant, i.e., as low as 0.5 wt. %, result in antiagglomeration. In low amounts, MeOH does not lead to salt deposition.51 It is also thought that MeOH will be present mostly in the bulk water phase and the aqueous side of the interfacial region in these mixtures. It appears that MeOH cosurfactant will be effective above a specific minimum concentration, and higher MeOH concentration will not be required. In our work, this concentration is likely around 0.5 wt. % or slightly less. Other alcohols may also be used as cosurfactants. According to Figure 8, the presence of MeOH cosurfactant does aid antiagglomeration via rhamnolipid but only down to certain rhamnolipid concentrations. At 0.1 wt. % rhamnolipid, significant adhesion occurs no matter how much MeOH is added. At 0.5 wt. % rhamnolipid, slurries exist in mixtures down to 0.5 wt. % MeOH. It was desired to determine if any rhamnolipid concentrations between these two values would also facilitate slurries. Thus, 0.25 wt. % rhamnolipid was also tested, and it is found that slurries are facilitated by this amount of surfactant but only when 2 or 5 wt. % MeOH cosurfactant is added. At 0.5 wt. % MeOH, mixtures with 0.25 wt. %
rhamnolipid show a tendency to allow significant hydrate adhesion upon vial walls. On the other hand, the quat is as effective as AA over all concentrations studied, i.e., down to 0.01 wt. %. A similar behavior is reported in our previous work and seems to indicate that the quat is effective at inducing steric repulsion between hydrate crystallites as well as in hydrate-wall interactions, whereas rhamnolipid may only be effective at both classes of repulsion when present in sufficient amount. The quat solution contains 15-20 wt. % 2-propanol, but it is assumed this does not play a cosurfactant role at the lower concentrations. For example, when 0.5 wt. % quat is added in mixtures of two parts isooctane, only ∼0.05 wt. %, with respect to the total water amount, of 2-propanol is present. When 0.01 wt. % quat is added, this amounts to only ∼2.3 × 10-3 wt. % of 2-propanol in the water phase. B. Kinetic/Thermodynamic Characteristics. Partitioning data in mixtures of four parts isooctane have been reported previously: the data reveal approximate aqueous phase THF concentrations of 10 wt. % at ambient temperature increasing to 15 wt. % at temperatures of 7 °C and below.21 We measured THF partitioning in the mixtures of two parts isooctane. The THF in the aqueous phase in these mixtures is about 15 wt. %, at ambient temperatures, to about 21 wt. % at lower temperatures. As shown in Figure 11 and Figure 13, Td values in mixtures without MeOH still range from 3.5 to roughly 4.3 °C, due to this partitioning of THF. Also shown in Figure 11 and Figure 13, a slight trend downward
10462 J. Phys. Chem. B, Vol. 112, No. 34, 2008
York and Firoozabadi
Figure 12. Average of freeze-thaw cycle data for mixtures of two parts isooctane, rhamnolipid, and low MeOH concentrations. Given are Tc (shown as 2), tc (shown as columns), and Td (shown as 9). Error bars are presented for all points; some Td and Tc error bars overlap and may not be clear.
Figure 13. Average of freeze-thaw cycle data for mixtures of two parts isooctane with or without ARQUAD 2C-75. Given are Tc (shown as 2), tc (shown as columns), and Td (shown as 9). Error bars are presented for all points; some may not be clear due to magnitude.
Figure 14. Average of freeze-thaw cycle data for mixtures of two parts isooctane, ARQUAD 2C-75, and low MeOH concentrations. Given are Tc (shown as 2), tc (shown as columns), and Td (shown as 9). Error bars are presented for all points; some Td and Tc error bars overlap and may not be clear.
in Tc values, i.e., a slight kinetic inhibition, is seen when surfactant is added in various amounts. This also means that ∆Ton increases when AA is added especially at lower concentrations. Figure 11 shows a clear trend for rhamnolipid mixtures; the same but less clear trend is seen in quat mixtures as given in Figure 13. The same trend has been observed in our previous work,39 yet the differences are not significant due to the scatter in the data. Figures 11 and 13 also reveal a decrease in dissociation temperature upon addition of AA, as expected from earlier work.21,39
Reliable crystallization and dissociation characteristics could be obtained by differential scanning calorimetry (DSC). In DSC, the sample size would be much smaller, on the order of 0.5 mL. Consequently, error in sample composition would be more difficult to prevent; also, DSC samples are not agitated so such a measurement would not be conducted upon an emulsified sample. The best practice is to conduct these measurements for crystallization and dissociation characteristics on the same samples and under the conditions used for agglomeration state assessment, the primary focus of these studies.
Alcohol Cosurfactants in Hydrate Antiagglomeration
Figure 15. A partial plug that appears more as a concentrated hydrate slurry in a mixture of zero parts isooctane, 1.5 wt. % rhamnolipid, and 10 wt. % MeOH. The steel ball is barely visible, but the air bubble shows the vial is not filled with a solid hydrate plug as would be expected. The vial in this image is tilted at roughly a 45° angle away from the borescope, as evidenced by position of bubble.
Another noticeable similarity between the current results, without MeOH, and our earlier work is that tc values decrease when rhamnolipid is added, yet this parameter increases when the quat is used. For all mixtures tested in this work, tc values are generally larger than the values in mixtures of four parts isooctane due to the increased amount of hydrate being formed when water-cut is larger. More hydrate is indeed being formed since the amount of water is larger and the THF amount is still in excess of the amount needed to convert all the water into hydrate. The theoretical hydrate volume fraction in such mixtures increases from 0.16 to 0.25 when the isooctane amount
J. Phys. Chem. B, Vol. 112, No. 34, 2008 10463 is decreased from four parts to two parts (see more detail in the caption to Figure 7). The effect of MeOH cosurfactant shown in Figures 12 and 14 is explained as follows. The effect of MeOH on Td appears as a clear trend in mixtures with rhamnolipid, Figure 12. That is, MeOH as a thermodynamic inhibitor depresses the equilibrium temperature, and Tc will also be reduced since it is generally less than Td through hysteresis, and the effect should be greater at higher concentrations. Taken in light of the scatter in the data, this trend is seen in Figure 12. This trend is not as clear for quat mixtures, shown in Figure 14, but in general Td values are lower when more MeOH is added. In both mixture types, values are in the general range expected, by both knowledge of thermodynamic inhibitors and from the only known published data54 on the effect of MeOH on THF hydrate equilibrium temperatures. The effect of MeOH cosurfactant on ∆Ton is unclear in either mixture type, largely due to the amount of scatter seen in the data. In rhamnolipid (Rh) mixtures, the data show increased ∆Ton values for the extremes in composition, i.e., for high Rh/ high MeOH and low Rh/low MeOH concentration, where high MeOH here simply means the higher concentration of 5 wt. %. However, Figure 12 clearly shows that a relatively high ∆Ton value also occurs for concentration of 0.5% Rh/2% MeOH. Figure 14 shows clearly a relatively large ∆Ton for composition of high quat/high MeOH, yet all other compositions exhibit a slim margin between Td and Tc values. The mixtures above exhibiting such a small ∆Ton value are not thought to be showing effects of hydrate promotion. Hydrate promotion is commonly shown through enhanced, i.e., increased, formation rate55,56 that is analogous to our crystallization time data. The MeOH cosurfactant does not appear to alter tc values significantly. As seen in Figures 3 and 4, the crystallization peaks in presence of MeOH are generally broader so this is likely offsetting the affect of increased driving force, i.e.,
TABLE 1: ARQUAD 2C-75-Induced Emulsion Stability Results (values in minutes) 2C-75 concentration (wt. %) 1.5 a
0.5
methanol (wt. %)
fresh
usedb
0 0.5 2 5
0.6 ( 0.1 0.4 ( 0.2 1.3 ( 0.4 0.6 ( 0.1
14.3 ( 0.6 17.0 ( 3.4 14.0 ( 1.0 22.3 ( 4.6
0.05
fresh
used
fresh
used
0.6 ( 0.2 0.3 ( 0.0 0.3 ( 0.1 0.3 ( 0.1
5.2 ( 2.0 2.7 ( 1.2 1.4 ( 0.9 10.7 ( 2.1
0.6 ( 0.0 0.3 ( 0.0 0.5 ( 0.0 0.4 ( 0.2
0.6 ( 0.1 0.2 ( 0.1 0.3 ( 0.1 7.0 ( 2.6
a Procedure used by Zanota, et al.21 A fresh sample, without stainless steel ball, is hand-agitated for 1 min and transferred at room temperature to a graduated cylinder and the time for separation of 60 volume % of the initial aqueous phase is measured and used as indicator of emulsion stability. b Samples used for visual observation and freeze-thaw cycles with stainless steel included are hand-agitated at room temperature for 1 min, and the time for separation of 60 volume % of the initial aqueous phase is measured and used as indicator of emulsion stability.
TABLE 2: Rhamnolipid-Induced Emulsion Stability Results (values in minutes) rhamnolipid concentration (wt. %) 1.5
0.5
0.05
methanol (wt. %)
fresha
usedb
fresh
used
fresh
used
0 0.5 2 5
0.4 ( 0.2 0.4 ( 0.1 0.8 ( 0.1 0.5 ( 0.2
2.3 ( 0.6 1.3 ( 0.6 2.3 ( 0.6 3.0 ( 0.9
0.5 ( 0.0 0.5 ( 0.1 0.6 ( 0.1 0.5 ( 0.2
2.1 ( 0.8 0.5 ( 0.0 1.2 ( 0.3 3.1 ( 1.6
0.2 ( 0.1 0.5 ( 0.3 0.2 ( 0.0 0.4 ( 0.1
0.3 ( 0.3 0.1 ( 0.1 0.2 ( 0.1 0.1 ( 0.0
a Procedure used by Zanota et al.21 A fresh sample, without stainless steel ball, is hand-agitated for 1 min and transferred at room temperature to a graduated cylinder and the time for separation of 60 volume % of the initial aqueous phase is measured and used as indicator of emulsion stability. b Samples used for visual observation and freeze-thaw cycles with stainless steel included are hand-agitated at room temperature for 1 min, and the time for separation of 60 volume % of the initial aqueous phase is measured and used as indicator of emulsion stability.
10464 J. Phys. Chem. B, Vol. 112, No. 34, 2008 supersaturation,57 at lower Tc. Only in about half the cases does the data show that addition of MeOH cosurfactant increases tc. Hydrate promotion in the presence of surfactant and MeOH would manifest itself through a significant decrease in tc values, which is not observed in our work. C. Emulsion Stability. Emulsion stability results in Tables 1 and 2 reveal the same effect in used samples, or those that have undergone freeze-thaw cycling, as seen in our previous work.39 However, these emulsions are mostly unstable, so the difference between the two test types, i.e., with fresh and used samples, is small. Stable emulsions in real pipeline fluids are undesirable,25,26 and therefore low stability values are acceptable. In general, the differences between the two tests are more significant for higher amounts of AA and MeOH cosurfactant. For fresh mixtures, there is little or no difference between stabilities of rhamnolipid or quat mixtures, with or without MeOH. There is some difference between rhamnolipid and quat mixtures for the used samples, but only at 1.5 and 0.5 wt. % AA. There appears to be no effect of MeOH cosurfactant on emulsion stability in fresh samples of either AA. The effect of MeOH is only noticeable in the used quat samples, with stability values jumping significantly upon addition of 5 wt. % MeOH for all concentrations of quat. For 1.5 and 0.5 wt. % quat, there is a downward trend in stability for lower MeOH concentrations; for 0.05 wt. % quat, stability values jump to roughly 7 min upon addition of 5 wt. % MeOH but drop to normally unstable values, i.e., on the order of 0.5 min or less, for smaller concentrations of MeOH. In used rhamnolipid samples, there are only slightly higher stabilities for higher MeOH amounts. For both AA mixtures with any amount of MeOH, it is clear from these results that stable emulsions are not necessary for effective antiagglomeration. 5. Conclusions In this work we have shown that small amounts of alcohol cosurfactant may have a significant effect on hydrate antiagglomeration. To our knowledge, this is the first report of alcohol cosurfactants in antiagglomeration of hydrates. We have shown the effectiveness of MeOH cosurfactant through visual observations with a multiple screening-tube rocking apparatus using high operating subcooling and residence time as indicators of performance. Shut-in and emulsion stability tests also lend supporting evidence. Several important conclusions can be drawn from our study. It is reconfirmed that rhamnolipid biosurfactants are attractive candidates for serious field testing. Another result reconfirmed here is that stable water-in-oil emulsions are not required for an effective AA. It is also again shown that AA may be effective at concentrations below what the conventional limit, i.e., 0.5 wt. %, is thought to be. Perhaps the most significant conclusion is that alcohol cosurfactants may be effective at low enough concentrations so that side-effects such as salt deposition from real fluids can be avoided. Experiments utilizing these ideas in this work are being planned in real fluids at high pressures. Acknowledgment. We are grateful to Jeneil Biosurfactant Co. for providing the rhamnolipid sample, as well as to Akzo Nobel for providing the quaternary ammonium salt sample. This work was supported by the member companies of the Reservoir Engineering Research Institute (RERI) in Palo Alto, CA. References and Notes (1) Sloan, E. D. Clathrate Hydrates of Natural Gases, 2nd ed.; Marcel Dekker: New York, 1998.
York and Firoozabadi (2) Dholabhai, P. D.; Parent, J. S.; Bishnoi, P. R. Fluid Phase Equilib. 1997, 141, 235. (3) Anderson, F. E.; Prausnitz, J. M. AIChE J. 1986, 32, 1321. (4) Christiansen, R. L.; Sloan, E. D. Ann. N. Y. Acad. Sci. 1994, 283. (5) Lederhos, J. P.; Long, J. P.; Sum, A.; Christiansen, R. L.; Sloan, E. D. Chem. Eng. Sci. 1996, 51, 1221. (6) Anklam, M. R.; Firoozabadi, A. J. Chem. Phys. 2005, 123, 144708. (7) Kelland, M. A.; Svartaas, T. M.; Dybvik, L. A. Proceedings of the SPE 69th Annual Technical Conference and Exhibition, New Orleans, LA, 1994. (8) Urdahl, O.; Lund, A.; Mørk, P.; Nilsen, T. N. Chem. Eng. Sci. 1995, 50, 863. (9) Huo, Z.; Freer, E.; Lamar, M.; Sannigrahi, B.; Knauss, D. M.; Sloan, E. D. Chem. Eng. Sci. 2001, 56, 4979. (10) Karaaslan, U.; Parlaktuna, M. Energy Fuels 2000, 14, 1103. (11) Arjmandi, M.; Ren, S. R.; Yang, J.; Tohidi, B. Proceedings of the Fourth International Conference on Gas Hydrates, Yokohama, Japan, 2002. (12) Kelland, M. A.; Svartaas, T. M.; Øvsthus, J.; Tomita, T.; Mizuta, K. Chem. Eng. Sci. 2006, 61, 4290. (13) Pakulski, M.; Hurd, D. Proceedings of the Fifth International Conference on Gas Hydrates, Trondheim, Norway, 2005. (14) Luckham, P. F. AdV. Colloid Interface Sci. 1991, 34, 191. (15) Israelachvili, J. N.; Wennerstro¨m, H. J. Phys. Chem. 1992, 96, 520. (16) Taylor, C. J.; Dieker, L. E.; Miller, K. T.; Koh, C. A.; Sloan, E. D. J. Colloid Interface Sci. 2007, 306, 225. (17) Kelland, M. A. Energy Fuels 2006, 20, 825. (18) Klomp, U. C.; Kruka, V. C.; Reijnhart, R. World Patent Application WO 9517579, 1995. (19) Sloan, E. D. Fluid Phase Equilib. 2005, 228-229, 67. (20) Urdahl, O.; Lund, A.; Mork, P.; Nilsen, T. N. Chem. Eng. Sci. 1995, 50, 863. (21) Zanota, M. L.; Dicharry, C.; Graciaa, A. Energy Fuels 2005, 19, 584. (22) Behar, P.; Kessel, D.; Sugier, A.; Thomas, A. Proceedings of the 70th Gas Processors Association Conference, San Antonio, TX, 1991. (23) Phillips, N. J.; Kelland, M. A. In Industrial Applications of Surfactants IV; Karsa, D. R., Ed.; Royal Society of Chemistry: London, 1999; pp 244. (24) Jakobsen, T.; Sjo¨blom, J.; Ruoff, P. Colloids Surf., A 1996, 112, 73. (25) Sjo¨blom, J.; Hemmingsen, P. V.; Kallevik, H. The Role of Asphaltenes in Stabilizing Water-in-Crude Oil Emulsions. In Asphaltenes, HeaVy Oils, and Petroleomics; Mullins, O. C., Sheu, E. Y., Hammami, A., Marshall, A. G., Eds.; Springer-Verlag: New York, NY, 2007; pp 549. (26) Førdedal, H.; Schildberg, Y.; Sjo¨blom, J.; Volle, J.-L. Colloids Surf., A 1996, 106, 33. (27) Andersson, V.; Gudmundsson, J. S. Ann. N. Y. Acad. Sci. 2000, 912, 322. Gas hydrates. (28) Alapati, R.; Lee, J.; Beard, D. Paper presented at the Offshore Technology Conference 2008, OTC 19505. (29) Lund, A.; Urdahl, O.; Kirkhorn, S. S. Chem. Eng. Sci. 1996, 51, 3449. (30) Fadnes, F. H. Fluid Phase Equilib. 1996, 117, 186. (31) Zhou, J. S.; Dupeyrat, M. J. Colloid Interface Sci. 1990, 134, 320. (32) Colafemmina, G.; Palazzo, G.; Balestrieri, E.; Giomini, M.; Giustini, M.; Ceglie, A. Prog. Colloid Polym. Sci. 1997, 105, 281. (33) Palazzo, G.; Lopez, F.; Giustini, M.; Colafemmina, G.; Ceglie, A. J. Phys. Chem. B 2003, 107, 1924. (34) Santhanalakshmi, J.; Parameswari, A. J. Surface Sci. Technol. 1997, 13, 123. (35) Sloan, E. D.; Subramanian, S.; Matthews, P. N.; Lederhos, J. P.; Khokhar, A. A. Ind. Eng. Chem. Res. 1998, 37, 3124. (36) Anklam, M. R.; York, J. D.; Helmerich, L.; Firoozabadi, A. AIChE J. 2008, 54, 565. (37) Kitamoto, D.; Yanagishita, H.; Endo, A.; Nakaiwa, M.; Nakane, T.; Akiya, T. Biotechnol. Prog. 2001, 17, 362. (38) Høiland, S.; Borgund, A. E.; Barth, T.; Fotland, P.; Askvik, K. M. Proceedings of the Fifth International Conference on Gas Hydrates, Trondheim, Norway, 2005. (39) York, J. D.; Firoozabadi, A. J. Phys. Chem. B 2008, 112, 845. (40) Kitamoto, D.; Isoda, H.; Nakahara, T. J. Biosci. Bioeng. 2002, 94, 187. (41) Fu, B.; Houston, C.; Spratt, T. Proceedings of the Fifth International Conference on Gas Hydrates, Trondheim, Norway, 2005. (42) Nitschke, M.; Costa, S. G. V. A. O.; Contiero, J. Biotechnol. Prog. 2005, 21, 1593. (43) Xie, Y.; Ye, R.; Liu, H. Colloids Surf., A 2007, 292, 189. (44) Hanley, H. J. M.; Meyers, G. J.; White, J. W.; Sloan, E. D. Int. J. Thermophys. 1989, 10, 903. (45) Makogon, T. Y.; Larsen, R.; Knight, C. A.; Sloan, E. D. J. Cryst. Growth 1997, 179, 258. (46) Devarakonda, S.; Groysman, A.; Myerson, A. S. J. Cryst. Growth 1999, 204, 525.
Alcohol Cosurfactants in Hydrate Antiagglomeration (47) Tohidi, B.; Anderson, R.; Clennell, M. B.; Burgass, R. W.; Biderkab, A. B. Geology 2001, 29, 867. (48) Lee, J. D.; Susilo, R.; Englezos, P. Chem. Eng. Sci. 2005, 60, 4203. (49) Zhang, Y.; Debenedetti, P. G.; Prud’homme, R. K.; Pethica, B. A. J. Phys. Chem. B 2004, 108, 16717. (50) Tombari, E.; Presto, S.; Salvetti, G.; Johari, G. P. J. Chem. Phys. 2006, 124, 154507. (51) Masoudi, R.; Tohidi, B.; Danesh, A.; Todd, A. C.; Yang, J. SPE Prod. Oper. 2006, 21, 182. (52) Usui, H.; Modak, P. R.; Suzuki, H.; Okuma, O. J. Chem. Eng. Jpn. 2004, 37, 15.
J. Phys. Chem. B, Vol. 112, No. 34, 2008 10465 (53) Inaba, H.; Inada, T.; Horibe, A.; Suzuki, H.; Usui, H. Int. J. Refrig. 2005, 28, 20. (54) Davidson, D. W.; Gough, S. R.; Ripmeester, J. A. Can. J. Chem. 1981, 59, 2587. (55) Zhong, Y.; Rodgers, R. E. Chem. Eng, Sci. 2000, 55, 4175. (56) Karaaslan, U.; Parlaktuna, M. Energy Fuels 2000, 14, 1103. (57) Kashchiev, D.; Firoozabadi, A. J. Cryst. Growth 2002, 241, 220. (58) Otake, K.; Tsuji, T.; Sato, I.; Akiya, T.; Sako, T.; Hongo, M. Fluid Phase Equilib. 2000, 171, 175.
JP8017265
