Synergistic Hydrate Inhibition of Monoethylene Glycol with Poly

Jul 7, 2014 - This study investigates the hydrate inhibition performance of monoethylene glycol (MEG) with poly(vinylcaprolactam) (PVCap) for retardin...
6 downloads 14 Views 2MB Size
Download PDF
Article pubs.acs.org/JPCB

Synergistic Hydrate Inhibition of Monoethylene Glycol with Poly(vinylcaprolactam) in Thermodynamically Underinhibited System Jakyung Kim,† Kyuchul Shin,† and Yutaek Seo* Division of Ocean Systems Engineering, Korea Advanced Institute of Science and Technology (KAIST), Daejeon 305-701, Republic of Korea

Seong Jun Cho and Ju Dong Lee Offshore Plant Resources R&D Center, Korea Institute of Industrial Technology, Busan 618-230, Republic of Korea S Supporting Information *

ABSTRACT: This study investigates the hydrate inhibition performance of monoethylene glycol (MEG) with poly(vinylcaprolactam) (PVCap) for retarding the hydrate onset as well as preventing the agglomeration of hydrate particles. A high-pressure autoclave was used to determine the hydrate onset time, subcooling temperature, hydrate fraction in the liquid phase, and torque changes during hydrate formation in pure water, 0.2 wt % PVCap solution, and 20 and 30 wt % MEG solutions. In comparison to water with no inhibitors, the addition of PVCap delays the hydrate onset time but cannot reduce the hydrate fraction, leading to a sharp increase in torque. The 20 and 30 wt % MEG solutions also delay the hydrate onset time slightly and reduce the hydrate fraction to 0.15. The addition of 0.2 wt % PVCap to the 20 wt % MEG solution, however, delays the hydrate onset time substantially, and the hydrate fraction was less than 0.19. The torque changes were negligible during the hydrate formation, suggesting the homogeneous dispersion of hydrate particles in the liquid phase. The well-dispersed hydrate particles do not agglomerate or deposit under stirring. Moreover, when 0.2 wt % PVCap was added to the 30 wt % MEG solution, no hydrate formation was observed for at least 24 h. These results suggest that mixing of MEG with a small amount of PVCap in underinhibited conditions will induce the synergistic inhibition of hydrate by delaying the hydrate onset time as well as preventing the agglomeration and deposition of hydrate particles. Decreasing the hydrate fraction in the liquid phase might be the reason for negligible torque changes during the hydrate formation in the 0.2 wt % PVCap and 20 wt % MEG solution. Simple structure II was confirmed by in situ Raman spectroscopy for the synergistic inhibition system, while coexisting structures I and II are observed in 0.2 wt % PVCap solution. approaches for avoiding hydrate risks in offshore flowlines have thus been moving from complete avoidance toward risk management, which involves allowing hydrate formation in flowlines but delaying nucleation or preventing agglomeration of hydrate particles using kinetic hydrate inhibitors (KHIs) or antiagglomerants (AAs), commonly called as low-dosage hydrate inhibitors (LDHIs).5,7 Since the kinetic inhibition effect of poly(vinylpyrrolidone) (PVP) and poly(vinylcaprolactam) (PVCap) was discovered, numerous water-soluble polymers such as antifreeze protein (AFP)-based polymers and polyaspartamides as well as PVPand PVCap-based polymers have been investigated and tested

1. INTRODUCTION Gas hydrates naturally occurring in seafloor sediments or permafrost have been highlighted as a potential source for methane gas;1−4 however, those in offshore flowlines have been a serious concern in the oil and gas industry because their formation can cause blockage leading to costly production stoppage and complex remediation work.5,6 To avoid hydrate formation in offshore flowlines transporting hydrocarbons, thermodynamic hydrate inhibitors (THIs) such as methanol and monoethylene glycol (MEG) have been commonly injected to shift the hydrate equilibrium curve outside the flowline operating conditions.6 However, this conventional method is now facing difficulties associated with larger injection volume, such as space limitations on floating structures and bulk logistics, as the energy industry moves to deeper and colder regions of the subsea to produce hydrocarbons. Recent © 2014 American Chemical Society

Received: April 8, 2014 Revised: June 27, 2014 Published: July 7, 2014 9065

dx.doi.org/10.1021/jp503435t | J. Phys. Chem. B 2014, 118, 9065−9075

The Journal of Physical Chemistry B

Article

for their inhibition performance.7−10 Although the inhibition mechanisms of KHIs is still unclear, recent molecular dynamics simulation studies suggest two roles of KHIs: (1) disrupting the organization of the forming clathrate by strong interactions with liquid water phase and (2) binding to the surface of hydrate crystal and retarding further growth once nucleation occurs.11−13 The primary purpose of KHI injection into flowlines is to delay hydrate onset and to avoid rapid crystal growth leading to serious plugging of flowlines until the hydrocarbon fluid arrives at the platform. However, the industry still relys on proven THIs for hydrate inhibition in the offshore oil and gas fields14 because KHIs may have limitations such as difficulty in preventing structure I (sI) hydrates15−17 and ineffectiveness at subcooling conditions higher than 14 °C, especially during the extended shut-in of flowlines.18 Although KHIs would reduce the growth rate of hydrate crystals by binding to growth sites, the hydrate fraction eventually will become high where the hydrate particles would deposit and block the hydrocarbon flow, as seen in the flowloop study.19 There have been attempts to develop synergist materials that improve the performance of KHIs. It has been found that poly(ethylene oxide) or glycol ether compounds noticeably prolong the nucleation time and extend the delay of catastrophic growth significantly;20,21 however, its effect is limited on delaying the hydrate onset. Therefore, the limitation of using KHI still remains. As the calculation of THI injection rate in offshore flowlines is based on the worst operational conditions with shut-in pressure, ambient seawater temperature, and maximum water production rate, the injection rate of THI used to be overestimated. The industry is making huge operating and capital expenditures for hydrate prevention via methanol or MEG injection;14 thus, even a small reduction of the THI injection rate can result in great savings. Here, we present the potential of synergistic hydrate inhibition in a thermodynamically underinhibited system by mixing MEG with PVCap, which is one of the most common KHIs. To the best of our knowledge, the hydrate inhibition performance when PVCap is added to MEG solution has not been investigated thoroughly and could offer quantitative estimation for reducing the amount of THIs. We studied hydrate formation characteristics, including hydrate onset time, subcooling temperature, hydrate fraction, and torque changes, for pure water, 0.2 wt % PVCap, 20 wt % MEG, 30 wt % MEG, 0.2 wt % PVCap + 20 wt % MEG, and 0.2 wt % PVCap + 30 wt % MEG solutions. Additionally, in situ Raman spectroscopic analysis was carried out to study the structural distribution of hydrates formed in the presence of both thermodynamic and kinetic hydrate inhibitors.

The pressure is measured by a pressure transducer with an uncertainty of 0.1 bar in a range of 0−200 bar. To provide vigorous mixing in the liquid phase, a four-blade mixer on a solid shaft coupled with the motor (BLDC 90) is used. The impeller is located on the base of the shaft, and the stirring rate is maintained at 600 rpm for all experiments. A torque sensor (TRD-10KC) with platinum-coated connector measures the torque of continuously rotating shaft with an uncertainty of 0.3%. It uses a strain gauge applied to a rotating shaft, and a slip ring supplies power to the strain gauge bridge. A picture of the autoclave setup is shown in Figure S1 of Supporting Information. The cell was immersed in a temperaturecontrolled liquid bath connected to an external refrigerated heater. After the cell was purged three times with the natural gas, the autoclave was pressurized to 120 bar at 24 °C while stirring at 600 rpm to saturate the aqueous phase with gas. Once the pressure and temperature reached steady-state, the cell was cooled to the desired temperature within an hour and kept for 24 h at that temperature. While the fluids were being cooled, pressure and temperature were continuously monitored. Three experiments were carried out for each system to average hydrate onset time, subcooling temperature, and the amount of consumed gas and to obtain reproducible trends for hydrate formation. A total of 24 experiments were carried out for pure water, 0.2 wt % PVCap solution, and 20 and 30 wt % MEG solutions at 3 and 5 °C. An additional 18 experiments were performed to investigate the synergistic inhibition of MEG with 0.2 wt % PVCap at 3, 4, and 5 °C. Figure 1 shows an

Figure 1. Gas consumption profiles during hydrate formation in natural gas and pure water system. Solid line represents the average of consumed gas values at corresponding time. The initial condition is 120 bar and 24 °C before commencing the cooling process. The temperature reaches 3 °C within an hour after commencing the continuous cooling; however, the hydrate onset is confirmed at 16.1 °C and 114.1 bar.

2. MATERIAL AND METHODS A synthetic natural gas which was composed of 90.0 mol % CH4, 6.0 mol % C2H6, 3.0 mol % C3H8, and 1.0 mol % n-C4H10 was supplied by Special Gas (Korea). Monoethylene glycol with 99.5% purity was purchased from Sigma-Aldrich. Deionized water and PVCap (MW ≈ 5000, purity 98.0 wt %) were used without further purification. A high-pressure autoclave made of SUS 316 was used in this study, which is equipped with magnet stirrer coupling and fourblade impeller. A total of 80 mL of the liquid was loaded in the autoclave cell with an internal volume of 360 mL. A platinum resistance thermocouple monitors the temperature of the liquid fluids inside of the autoclave with an uncertainty of 0.15 °C.

example of gas consumption profiles during the hydrate formation in natural gas and pure water system. Time zero indicates the moment of hydrate onset, and from that time the consumption of gas can be calculated from the pressure difference between the measurement moment and calculated pressure assuming no hydrate was formed. As can be seen in Figure 1, the gas consumption profiles show trends similar to each other, representing the standard deviation of 6% with the maximum deviation of 31%. The 9066

dx.doi.org/10.1021/jp503435t | J. Phys. Chem. B 2014, 118, 9065−9075

The Journal of Physical Chemistry B

Article

Table 1. Equilibrium Temperature, Hydrate Onset Temperature, Subcooling Temperature, Hydrate Onset Time, and Hydrate Fraction for Pure Water, PVCap 0.2 wt %, MEG 20 wt %, and MEG 30 wt % Solutions at Target Cooling Temperature of 3.0 °C system

Peq (bar)

Teq (°C)

Tonset (°C)

ΔTsub (°C)

tonset (min)

Φhyd

τini (N cm)

τmax (N cm)

τsteady (N cm)

pure water PVCap 0.2 wt % MEG 20 wt % MEG 30 wt %

119.1 117.9 113.2 112.2

21.7 21.6 15.8 11.6

16.1 12.5 9.9 5.5

5.6 9.1 5.9 6.1

19.8 33.8 22.8 27.3

0.75 0.65 0.28 0.17

4.7 4.5 4.8 4.6

35.2 19.1 5.8 5.8

11.2 14.2 4.9 4.9

calculated pressure with the assumption that no hydrate was formed. This procedure has been suggested in the literature24 as a method for studying hydrate formation in a flow wheel apparatus. Thus

experimental apparatus and procedure in this work provide reproducible results for studying hydrate formation characteristics. The performance of hydrate inhibitor in preventing agglomeration of hydrate particles was evaluated using torque changes because the torque rises sharply because of restriction on the overhead stirrer when agglomeration and deposition of hydrate particles occur. Detailed description of the in situ Raman and 13C NMR analysis are presented in our previous works.17,22,23 The experimental apparatus for in situ Raman analysis mainly consists of a stainless-steel reactor, syringe pump, refrigeration system, water bath, data acquisition system, and real-time Raman spectroscopy instrument. A Sentinel Raman spectrometer (Bruker Optics Ltd.) with a Unilab II probe (fiber optic) and a CCD (charge coupled device) detector was used. In situ Raman analysis was carried out using optical fibers, allowing a Raman probe tip to be coupled to a spectrometer. A single cable is used to transmit the laser to the sample; at the same time, another fiber is then used to transfer the Raman signal from the sample to a standard spectrometer and detection system. These two cables are connected to a compact, rugged Raman probe tip that focuses the laser onto the sample and collects the Raman signals. The probe tip was especially designed and inserted into the hydrate reactor for in situ Raman analysis. A 9.5 mm diameter cylindrical stainless-steel probe tip with an 8 mm ball lens (N-BK7m Edmund optics) was inserted through a fitting in the wall of the reactor. The front focal length (FFD = 3.7 mm) was determined by the “lens-maker’s formula” using the refractive index and radius of curvature of the surfaces of the lens. A 532 nm Nd:Yag laser source with an incident laser power of 100 mV was used for excitation. The spectral range was 500−4400 cm−1, and the spectral resolution was 4−6 cm−1. During the hydrate formation, Raman spectra of two systems, which are 0.2 wt % PVCap only and 20 wt % MEG + 0.2 wt % PVCap solutions, were recorded every 6 min with a 3 min exposure time. The pressure of the reactor was kept constant at 70 bar during the entire experiment by a high-pressure syringe pump (Teledyne, ISCO 500D), and the constant cooling method from 24 to 3 °C within 1 h was used for the hydrate formation. When the formation process was completed, as confirmed by steady volume in the syringe pump, the system was depressurized and then hydrate samples were transferred to the sample tube for 13 C NMR analysis. The experimental procedures and NMR spectrometer are explained in our previous works.17,23

ΔnH, t =

⎛ PexpVcell ⎞ ⎛ PcalVcell ⎞ ⎜ ⎟ − ⎜ ⎟ ⎝ zRT ⎠t ⎝ zRT ⎠ t

where ΔnH,t is the moles of gas consumed for hydrate formation at a given time, Pcal the calculated pressure assuming no hydrate, Pexp the measured pressure, Vcell the volume of gas phase, and z the compressibility factor calculated using Peng− Robinson equation of state. The hydrate fraction (ϕhyd) in the liquid phase is calculated using the hydration number of 6.5 and the following equation: ϕhyd =

Vhyd Vhyd + (Vw − Vw,conv)

where Vw is the volume of water, Vw,conv the volume of the water converted to hydrate, and Vhyd the volume of hydrate calculated from the molecular weight and density of hydrates calculated at a given time. The hydration number, 6.5, was obtained from cage occupancies of gas molecules in 512 and 51264 cages of structure II in our previous work.23 Recent studies on the formation of a hydrate plug using a flow loop and autoclaves suggest that hydrate fraction in the liquid phase is an important parameter to describe the flowability of hydrate particles.19,25 Hydrate particles are dispersed homogeneously in the liquid phase at low fraction but soon are segregated heterogeneously followed by deposition and bedding on the wall with increasing hydrate fraction. Although the flow parameters such as gas void fraction, gas density, and flow velocities may affect the deposition process of hydrate particles, the hydrate fraction in the liquid phase is central to understanding the agglomeration and deposition of hydrate particles. In this work, torque measurements were carried out to investigate the agglomeration and deposition of hydrate particles as a function of hydrate fraction in the liquid phase. The hydrate equilibrium temperature, hydrate onset temperature, subcooling temperature, hydrate onset time, and hydrate fraction are presented in Table 1. The three torque values are also shown in Table 1 to describe the torque changes with time. Typical torque change can be expressed by the initial value, the maximum value due to agglomeration and deposition of hydrate particles, and the steady values shown in the later stage of hydrate formation. Figure 2a,b shows the gas consumption profiles with time and torque changes as a function of hydrate fraction in the liquid phase when the target cooling temperature is 3 °C. The average hydrate onset time was 19.8 min, and the average subcooling temperature was 5.6 °C when hydrate forms from the pure water. The hydrate fraction in the liquid phase was in the range between 0.72 and 0.80, leading to an average

3. RESULTS AND DISCUSSION We first investigated the formation kinetics of hydrate in pure water, 0.2 wt % PVCap solution, and 20 and 30 wt % MEG solutions. Hydrate onset time was defined as the time elapsed from when the system enters the hydrate stability zone to the first indication of hydrate formation. The moles of gas consumed during hydrate formation were calculated from the pressure difference between the measurement moment and 9067

dx.doi.org/10.1021/jp503435t | J. Phys. Chem. B 2014, 118, 9065−9075

The Journal of Physical Chemistry B

Article

growth rate of hydrate. When 17.5 min passes since the hydrate onset, an inflection point occurs in the gas consumption curve and the gas consumption rate becomes much less than that of pure water. It takes 486 min for the 0.2 wt % PVCap solution gas consumption curve to reach a plateau at 0.43, whereas it takes 205 min for pure water to reach 0.54. The hydrate fraction is low compared to that of pure water; however, the torque rises earlier when the hydrate fraction reaches 0.11. Its value becomes close to that of pure water in the later stage, indicating the possibility of agglomeration and deposition of hydrate particles. The torque remains at 14 N cm with slight fluctuation even until the fraction increases to 0.60. It is worth noting that PVCap is an effective KHI, delaying the onset time and lowering the growth rate, but it cannot control the agglomeration and deposition of hydrate particles once the hydrate fraction is higher than 0.11. Dramatic differences in torque change and hydrate fraction were observed in underinhibited systems, 20 wt % MEG and 30 wt % MEG solutions. The initial pressure and temperature were set at 120 bar and 24 °C before commencing cooling to the target temperature. Considering the target temperature of 3 °C and the initial pressure of 120 bar, maintaining the MEG concentration in liquid water above 43.0 wt % is required to completely avoid the formation of hydrates. Once the concentration of MEG decreases to 20 wt %, the target cooling temperature of 3.0 °C will lead to the system falling into the hydrate stability zone, although the addition of 20 wt % MEG to pure water shifts the hydrate equilibrium curve. Table 1 presents the onset time of 22.8 min for 20 wt % MEG solution and 27.3 min for 30.0 wt % MEG solution. The weak kinetic inhibition performance is observed for the underinhibited system at the studied pressure and temperature conditions. However, as shown in Figure 2, the gas consumption reaches steady-state much earlier (only 22 min after the hydrate onset) than pure water and 0.2 wt % PVCap solution. The hydrate fraction reaches only 0.28 for 20 wt % MEG solution and 0.17 for 30 wt % MEG solution at the end of the experiments, which indicates the underinhibition system is able to reduce the hydrate fraction substantially, possibly due to self-inhibition effect in which the MEG concentration in the aqueous phase becomes higher while converting liquid water into hydrate crystals. It is noted that the converted amount of water to hydrate was 75% and 65% for pure water and 0.2 wt % PVCap solution, respectively, but it was only 28% and 17% for 20 wt % MEG and 30 wt % MEG solutions, respectively. When considering the fact that MEG will not be incorporated into hydrate cages and remain in free water, the concentration of MEG in liquid water phase increases during hydrate formation. From the calculated amount of water remaining as free water, the MEG concentration at the end of the experiment reaches 24.8 wt % for 20 wt % MEG solution and reaches 33.9 wt % for 30 wt % MEG solution. An increase of 4−5 wt % in MEG concentration is observed during the hydrate formation. Figure S2 of Supporting Information shows the PT-trace curves for pure water, 0.2 wt % PVCap, 20 wt % MEG, and 30 wt % MEG solutions. A slight increase in torque was observed upon hydrate onset; however, this quickly becomes close to the initial value and remains stable, indicating the hydrate particles were welldispersed in the liquid phase. In the case of deposition of hydrate particles on the wall, the torque would rise sharply and the stirrer may stop rotating because of restriction. However,

Figure 2. (a) Gas consumption profiles with time during hydrate formation at 3 °C. (b) Typical torque changes observed during hydrate formation in the presence of hydrate inhibitors at 3.0 °C. From top to bottom: pure water (no inhibitor), PVCap 0.2 wt %, MEG 20 wt %, and MEG 30 wt %.

value of 0.75 with the standard deviation of 0.05. Its relative standard deviation is 6.2%. It is expected that hydrates form and grow into particles that eventually agglomerate and deposit on the wall. There was no instant increase of torque upon the hydrate onset, but when the hydrate fraction reaches 0.29, the torque rises gradually leading to high torque values of 18.5 N cm with instant maximum torque of 35.2 N cm while the stirrer keeps rotating. The gradual increase of torque was observed 19.0 min after the hydrate onset. Thus, the catastrophic growth was observed 38.8 min after the temperature becomes lower than the hydrate equilibrium temperature. As shown in Table 1 and Figure 2, the addition of 0.2 wt % PVCap to pure water delays the hydrate onset and reduces the 9068

dx.doi.org/10.1021/jp503435t | J. Phys. Chem. B 2014, 118, 9065−9075

The Journal of Physical Chemistry B

Article

water system showed a typical hydrate formation, which is rapid gas consumption in the initial stage and reduced consumption followed by plateau in the later stage. The PVCap-added system, on the other hand, showed kinetically inhibited hydrate formation, relatively delayed hydrate onset time, and long propagation time of hydrate crystals. The two underinhibited systems, 20 wt % MEG and 30 wt % MEG solutions, initially showed rapid hydrate growth although the thermal driving force is reduced;26 however, the amount of gas consumption was quite small because of the self-inhibition effect. From the calculated amount of water remaining as free water, the MEG concentration at the end of the experiment reaches 25.1 wt % for 20 wt % MEG solution and reaches 33.4 wt % for 30 wt % MEG solution. A 3−5 wt % increase in MEG concentration is observed during the hydrate formation. The obtained results for MEG concentration increase may suggest the self-inhibition effect is substantial in suppressing the further growth of hydrate particles. Further experiments will be carried out to investigate the change of MEG concentration more precisely. Among the studied inhibition systems, 30 wt % MEG solution shows the best hydrate inhibition performance by delaying hydrate onset time and by maintaining a stable torque. However, it seems the inhibition performance is achieved mainly because of the prevention of hydrate particle agglomeration rather than delaying the hydrate onset time. Figure 3b shows the torque changes with hydrate fraction when the target cooling temperature was 5 °C, in which less thermal driving force is applied for the hydrate formation than at 3 °C. Once again, it is confirmed that the torque rises gradually when hydrate fraction reaches 0.26 for pure water and 0.14 for 0.2 wt % PVCap solution. The final hydrate fraction is 0.77 and 0.66 for pure water and 0.2 wt % PVCap solution, respectively. However, the underinhibited systems, 20 and 30 wt % MEG solutions, show no significant torque changes for all experiments, and hydrate fractions was 0.29 and 0.15, respectively. MEG is the most widespread thermodynamic hydrate inhibitor for offshore gas fields, and we selected MEG as a base hydrate inhibitor for investigating the synergist in this work. Our previous work27 suggested the addition of 1.0 wt % PVP to 30.0 wt % MEG solution incurs a slight delay in the hydrate onset time. This result leads us to investigate further the synergistic inhibition of PVCap and MEG using the abovedescribed constant cooling method. Although we maintain the temperature at 3 °C for 24 h with the 0.2 wt % PVCap and 30.0 wt % MEG solution, the pressure stays at 104.3 bar and there is no pressure drop due to hydrate formation. In addition, no temperature spike is observed. It is clear that the hydrate nucleation was delayed for more than 24 h in 0.2 wt % PVCap and 30 wt % MEG solution. In the above results, the hydrate formation was well-observed in the presence of PVCap only and in the underinhibited MEG solutions, as shown in Figures 2 and 3. When 0.2 wt % PVCap was added to 30 wt % MEG solution, no hydrate formation was observed for more than 24 h. This result indicates the strong kinetic inhibition performance is achieved by adding only 0.2 wt % PVCap to 30 wt % MEG solution. To provide more thermal driving force for hydrate formation, we decrease the temperature down to −2.0 °C, and finally we observe the hydrate formation from the pressure decrease from 104.3 to 93.4 bar that results from the hydrate fraction of 0.28 in the liquid phase. Figure 4 shows the pressure and temperature changes during the experiment and the observed

the underinhibited MEG system was effective in preventing the agglomeration and deposition of hydrate particles. Figure 3 shows the hydrate formation characteristics when the target cooling temperature is 5 °C. Hydrate onset time,

Figure 3. (a) Gas consumption profiles with time during hydrate formation at 5.0 °C. (b) Typical torque changes observed during hydrate formation in the presence of hydrate inhibitors at 5.0 °C. From top to bottom: pure water (no inhibitor), PVCap 0.2 wt %, MEG 20 wt %, and MEG 30 wt %.

subcooling temperature, and hydrate fraction are presented in Table 2. It is noted that there was only a slight difference between the results obtained at 3 and 5 °C. This is because the hydrate formation happens before the temperature reaches the target temperature. The subcooling temperatures in Table 2 suggest the hydrate formation occurs when the temperature is 4.6−8.9 °C below hydrate equilibrium temperature. The pure 9069

dx.doi.org/10.1021/jp503435t | J. Phys. Chem. B 2014, 118, 9065−9075

The Journal of Physical Chemistry B

Article

Table 2. Equilibrium Temperature, Hydrate Onset Temperature, Subcooling Temperature, Hydrate Onset Time, and Hydrate Fraction for Pure Water, PVCap 0.2 wt %, MEG 20 wt %, and MEG 30 wt % Solutions at Target Cooling Temperature of 5.0 °C system

Peq (bar)

Teq (°C)

Tonset (°C)

ΔTsub (°C)

tonset (min)

Φhyd

τini (N cm)

τmax (N cm)

τsteady (N cm)

pure water PVCap 0.2 wt % MEG 20 wt % MEG 30 wt %

117.7 118.7 114.1 112.1

21.6 21.7 15.8 11.5

17.0 12.8 10.5 5.7

4.6 8.9 5.3 5.8

15.7 33.0 20.8 26.0

0.77 0.66 0.29 0.15

4.6 4.4 4.7 4.8

19.1 18.7 5.7 5.5

9.8 8.4 5.1 4.7

Figure 4. Trace of pressure and temperature changes of 30 wt % MEG + 0.2 wt % PVCap system during temperature ramping process. The amount of consumed gas and hydrate fraction is calculated from the pressure measurement.

gas consumption. Once the hydrate fraction is confirmed to be close to 0.28, we stop stirring and then increase the temperature to 13.1 °C to dissociate the hydrates at the corresponding pressure. We confirm the complete dissociation of hydrate visually and from the stable pressure at 109.7 bar. After the temperature is maintained at 13.1 °C for 5.0 h, it decreases again back to 3 °C and is held for 24 h to observe the formation of hydrates. While the temperature is lowered to 3 °C, the hydrate formation is observed instantly; however, the hydrate fraction maintains at around 0.03. It is presumed that the remaining residual hydrate structures form hydrate instantly when the temperature decreases below the hydrate equilibrium temperature. Once the hydrate nucleus occurs, hydrate crystals used to keep growing continuously even though THI or KHI is present, as shown in Figures 2 and 3. However, in the presence of 30 wt % MEG and 0.2 wt % PVCap together, there was no further hydrate growth at 3 °C even with low hydrate fraction in the liquid phase (Figure 4). This implies that the synergy between MEG and PVCap significantly suppresses the hydrate growth as well as the nucleation. Therefore, the concept of synergistic

inhibition might be a new candidate for minimizing the risk of hydrate formation in offshore flowlines. To investigate the effect of changing MEG concentration on the synergistic inhibition, we studied the hydrate formation in 0.2 wt % PVCap and 20 wt % MEG solution. In Figure 5, all three repetitions of hydrate formation at (a) 3 °C, (b) 4 °C, and (c) 5 °C are shown. Hydrate onset time, subcooling temperature, and hydrate fraction are shown in Table 3. Two remarkable observations based on Figure 5 are that (1) the hydrate formation occurs stepwise and (2) changing the target cooling temperature only 1 °C drastically reduces the amount of hydrate formed. It is impossible to generate the average curve that was carried out in Figure 1. In Figure 5, we found that there exist at least two or three different hydrate formation steps and each step occurs stochastically after the onset of the former step. This stepwise formation of hydrates will be explored further in the near future. In Figure 5c, hydrate onset is observed 112.0 min after the temperature is lower than the hydrate equilibrium temperature. However, its growth is suppressed significantly, and the hydrate fraction remain less than 0.006 for 1000 min in the third experiment. The 9070

dx.doi.org/10.1021/jp503435t | J. Phys. Chem. B 2014, 118, 9065−9075

The Journal of Physical Chemistry B

Article

than the hydrate fraction in other experiments. This slow growth followed by catastrophic growth has been reported by many researchers as evidence of the role of PVCap acting on the growing surface of hydrate crystals.17 The results obtained in these experiments clearly indicate the kinetic hydrate inhibition performance is enhanced substantially when adding 0.2 wt % PVCap to 20 wt % MEG solution. As described above, the underinhibited MEG system was effective in preventing the agglomeration and deposition of hydrate particles, as shown in the low torque values in Tables 1 and 2. This feature is also observed for the 0.2 wt % PVCap and 20 wt % MEG solution. The torque remains stable during the formation of hydrate, and the hydrate fraction reaches only 0.20, which is close to the lowest hydrate fraction, 0.17, obtained from 30.0 wt % MEG solution when the target cooling temperature was set to 3.0 °C (Figure 6). The torque change

Figure 6. Torque changes as a function of hydrate fraction during hydrate formation in 20 wt % MEG + 0.2 wt % PVCap system at (a) 3 °C, (b) 4 °C, and (c) 5 °C.

with hydrate fraction in these results provides strong evidence that mixing of MEG with PVCap is effective for preventing the agglomeration and deposition of hydrate particles, which is different than surfactant-based antiagglomerant. This is an important finding because the mixing of chemical additives sometimes induces an adverse effect on the performance of each additive.18 However, the addition of 0.2 wt % PVCap to 20 wt % MEG solution enhances the kinetic inhibition performance and prevents the deposition of hydrate particles simultaneously; thus, this system shows true synergistic inhibition performance. A molecule of MEG has two hydroxyl groups forming hydrogen bonds with water

Figure 5. Gas consumption profiles with time during hydrate formation in 20 wt % MEG + 0.2 wt % PVCap system at (a) 3 °C, (b) 4 °C, and (c) 5 °C.

catastrophic growth is observed at 1000 min after the hydrate onset, but the hydrate fraction reaches 0.05, which is still less

Table 3. Equilibrium Temperature, Hydrate Onset Temperature, Subcooling Temperature, Hydrate Onset Time, and Hydrate Fraction for 0.2 wt % PVCap and MEG 20 wt % Solution When Varying Target Cooling Temperature temperature (°C)

Peq (bar)

Teq (°C)

Tonset (°C)

ΔTsub (°C)

tonset (min)

Φhyd

τini (N cm)

τmax (N cm)

τsteady (N cm)

3 4 5

115.0 117.8 112.4

15.9 16.0 15.8

3.2 4.1 4.9

12.7 11.9 10.9

111.0 110.7 112.0

0.19 0.06 0.02

4.9 4.5 4.7

6.1 6.0 5.8

5.1 5.3 5.1

9071

dx.doi.org/10.1021/jp503435t | J. Phys. Chem. B 2014, 118, 9065−9075

The Journal of Physical Chemistry B

Article

molecules similar to that of hydrate formation, which is called a clustering effect. Thus, it reduces water activity, shifting the hydrate equilibrium conditions to lower temperature and higher pressure. As the driving force for the intrinsic growth of hydrate used to be expressed by the kinetic constants and the temperature difference between the hydrate equilibrium condition and the system operation condition,29 the concentration increase of MEG during hydrate formation will reduce the driving force for the hydrate growth. Furthermore, as discussed in the literature,12 PVCap molecules disrupt the local organization of the water and gas molecules, increasing the barrier to nucleation. Once nucleation occurs, PVCap binds to the surface of the hydrate crystal and retards further growth along the growth plane. The nucleation and growth model30 adopting the classical crystallization theory represents the expression using the kinetic parameter depending on the mechanism of attachment of hydrate building units to the hydrate nucleus. The binding of PVCap to the hydrate crystal will reduce the number of nucleation sites. Therefore, in the presence of MEG and PVCap in the aqueous phase, the thermodynamic-dependent driving force and the intrinsic kinetics rate will be reduced simultaneously during the formation of hydrates. Once again, it is noted that PVCap is able to associate well with the MEG; consequently both additives perform synergistically to inhibit the nucleation and growth of hydrates. In previous works15,16,18 it was suggested that a kinetic hydrate inhibitor such as PVCap is more effective for inhibiting the nucleation and growth of sII hydrates. It may accompany the formation of metastable sI hydrates even though the gas composition thermodynamically prefers to form sII hydrates. This structural distribution during hydrate formation may increase the complexity of managing hydrate formation and dissociation in offshore flowlines. The in situ Raman spectroscopic study was carried out to investigate the structural characteristics of the hydrates formed from 0.2 wt % PVCap and 20 wt % MEG solution. The in situ Raman spectroscopic observation for the region of methane C−H stretching provides evidence on the hydrate structures during the formation process. In Figure 7a, the Raman spectra of 0.2 wt % PVCap only system are shown. The one peak at 2909 cm−1 before hydrate onset (−6 min) comes from dissolved methane in the liquid phase.28 The peak indicating methane molecules encaged in the 512 cages (small cages) appears at 2914 cm−1 upon hydrate onset (0 min) and then grows further. The peak for methane molecules in the 51264 cages (large cages) appears at 2902 cm−1 at 6 min after the hydrate onset. It is unclear whether there is only one hydrate structure from the Raman peaks position; thus, we decided to calculate the ratio of intensity for methane peaks that will provide important information on the structural characteristics of formed hydrates. When forming hydrates with natural gas, the preferable structure is sII because of the presence of ethane and propane molecules. As they are enclathrated only in large cages of sII, methane molecules used to occupy small parts of the large cages but most of the small cages. The methane peaks for the molecules in the 51262 cages and the 51264 cages are not distinguishable from each other in the obtained Raman spectra because of the limitation of the instrument. However, it is possible to identify the occurrence of sI from the ratio of the peaks at 2902 and 2914 cm−1. At 6 min after the onset, the peak for methane in small cages grows further and the peak for methane in large cages appears with small intensity, leading to

Figure 7. Raman spectra of (a) 0.2 wt % PVCap and (b) 20 wt % MEG + 0.2 wt % PVCap solutions. (c) Ratio of intensities between the peaks at 2912 and 2902 cm−1. The time when the methane peak in the hydrate was first observed was set to 0 min.

the intensity ratio of the two peaks to be 5.1. At 12 min after the onset, the intensity of peak at 2902 cm−1 suddenly increases and the ratio of Ismall (2914 cm−1) to Ilarge (2902 cm−1) 9072

dx.doi.org/10.1021/jp503435t | J. Phys. Chem. B 2014, 118, 9065−9075

The Journal of Physical Chemistry B

Article

Figure 8. 13C NMR spectra of hydrate samples formed from (a) 0.2 wt % PVCap solution and (b) 0.2 wt % PVCap and 20 wt % MEG solution. CH4 molecules in hydrate cages appear in the chemical shift range between 0 and −10 ppm, while C2H6 molecules in hydrate cages appear in the chemical shift rage between 10 and 0 ppm.

from 13C NMR spectra in Figure 8a. On the other hand, for the case of 0.2 wt % PVCap and 20 wt % MEG solution, typical growth of sII is observed in Figure 7b and the 13C NMR spectra in Figure 8b confirms the sII hydrate with the intensity ratio of 4.0. This spectroscopic analysis suggests that the occurrence of metastable sI was prevented in the presence of both PVCap and MEG. The industrial practice to prevent hydrate-related risks in offshore flowlines has been injecting thermodynamic hydrate inhibitors such as MEG. Alternative materials exist for hydrate prevention, including kinetic hydrate inhibitors and antiagglomerants. Efforts have focused on developing novel polymer structures or surfactants to improve the kinetic inhibition or antiagglomeration performance. In this study, the effect of adding a small amount of PVCap to MEG solution on the hydrate formation characteristic was investigated as a new method for managing the hydrate formation. For PVCap alone system, weak kinetic inhibition performance was observed, but when PVCap and MEG are combined, i.e., kinetic hydrate inhibitor and the underinhibited MEG solution, the kinetic inhibition performance was improved substantially (about 3.4 times). Moreover, as seen in Figure 5c, the long period of slow hydrate growth before catastrophic growth will provide additional benefit for managing hydrate formation. In addition to the improved kinetic inhibition performance, the hydrate fraction in the liquid phase remains below 0.19. Recent studies from flow loop tests and accompanied simulation suggest that hydrate deposition in offshore flowlines could occur for hydrate volume fractions greater than 0.3, while hydrate with fraction between 0.1 and 0.3 could be safely transported in the liquid phase.24,25 It can be suggested that by controlling the concentration of MEG the hydrate fraction in the liquid phase can be reduced using the self-inhibition effect of MEG with the synergistic inhibition concept. Additional benefit of using the mixture of PVCap and MEG was suppressing the formation of metastable sI hydrate, thus

decreases from 5.1 to 1.7 until 66 min after the onset (Figure 7c). On the other hand, the Raman spectra of hydrates formed from 0.2 wt % PVCap + 20 wt % MEG solution in Figure 7b show relatively slow growth of both peaks at 2914 and 2902 cm−1, and the intensity ratio shown in Figure 7c gradually decreases and approaches ∼4.0 when 570 min passes after the hydrate onset. Panels a and b of Figure 8 show the 13C NMR spectra obtained from the hydrate samples in the presence of 0.2 wt % PVCap only and 0.2 wt % PVCap + 20 wt % MEG solution, respectively. The methane molecules show two resonances at −4.3 and −8.3 ppm when they occupy the small and large cages of sII, respectively. An additional resonance in Figure 8a at −6.8 ppm is from methane molecules occupying the large cages of sI.17,23 Therefore, it is confirmed that only sII has been formed from 0.2 wt % PVCap + 20 wt % MEG solution; however, both sI and sII coexist when hydrate forms from 0.2 wt % PVCap solution. In Figure 8a,b, single resonance for ethane is confirmed at 6.1 ppm, indicating ethane molecules occupy only the large cages of structure II. It is noted that they do not participate in the formation of structure I when metastable structure I has formed in 0.2 wt % PVCap solution. The intensity ratio of methane in small to large cages of sII hydrate was 4.0 for 0.2 wt % PVCap + 20 wt % MEG solution. This is because sII hydrate has a greater number of small cages in the unit cell (16 small and 8 large cages per 136 waters) than sI hydrate (2 small and 6 large cages per 46 waters) and because a methane molecule prefers the most to be encaged in 512 small cages rather than in 51264 large cage. When both sI and sII exist together, as is the case of hydrate formed from 0.2 wt % PVCap solution (Figure 8a), the intensity ratio of methane in small cages (512) to large cages (51262 and 51264) for both sI and sII was 1.8. Both intensity ratios from NMR and Raman analysis are in good agreement, which implies the sudden appearance of metastable sI at 12 min after the hydrate onset, as shown in Figure 7a. Two coexisting sI and sII hydrates are confirmed 9073

dx.doi.org/10.1021/jp503435t | J. Phys. Chem. B 2014, 118, 9065−9075

The Journal of Physical Chemistry B

Article

observed for 0.2 wt % PVCap solution, which indicates that the formation of metastable structure I can be suppressed in the presence of MEG. These results suggest that mixing of a small amount of PVCap with MEG in underinhibited conditions will induce the synergistic inhibition of hydrate by delaying the hydrate onset time as well as preventing the agglomeration and deposition of hydrate particles. It is worth noting that the MEG concentration can be reduced substantially by adopting the synergistic inhibition concept instead of using the thermodynamic hydrate inhibitor; thus, we believe this is a promising alternative to the typical hydrate inhibition strategy.

minimizing the complexity of handling different hydrate structures in managing the hydrate risk. The industry has been relying on the injection of overestimated MEG; thus, its reduction can lead to great cost savings for not only the injected amount of MEG but also the facilities issues such as topside space limitations, MEG distillation efficiency, and bulk logistics issues. When transporting the natural gas under a pressure of 100 bar with an ambient temperature of 4 °C, the MEG concentration required to completely prevent hydrate formation thermodynamically in offshore flowline is 43 wt %. The obtained results in this work suggest that the injection rate of MEG can be reduced from 43 wt % to 20 wt % when choosing the synergistic inhibition concept of adding 0.2 wt % PVCap to 20 wt % MEG solution. Other variables, such as the gas−water ratio, salt concentrations, flowline geometry, and flow velocities, are also important for determining the characteristics of hydrate formation and its transportation in offshore flowlines. However, assessing the formation characteristics using both hydrate onset time and the hydrate fraction can be a good starting point for determining the hydrate risk in operation of offshore flowlines. The synergistic hydrate inhibition using a small amount of PVCap and underinhibited MEG increases the hydrate onset time substantially and prevents the agglomeration of hydrate particles; thus, we believe this is a promising alternative to the typical thermodynamic hydrate inhibitor-based strategy. Additional laboratory experiments will be carried out with the addition of light condensate, in the presence of salt, and with other spectroscopic analyses.

■

ASSOCIATED CONTENT

S Supporting Information *

Image of the experimental apparatus and the pressure− temperature trace curve details. This material is available free of charge via the Internet at http://pubs.acs.org

■

AUTHOR INFORMATION

Corresponding Author

*Tel: +82 42 350 1521. Fax: +82 42 350 1510. E-mail: Yutaek. [email protected]. Author Contributions †

J.K. and K.S. contributed equally to this work.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Global Leading Technology Program of the Office of Strategic R&D Planning (OSP) funded by the Ministry of Trade Industry & Energy (10042424) and by the Industrial Infrastructure Program (Infrastructure for Offshore Plant Resources R&D Center) through the Korea Institute for Advancement of Technology (KIAT) grant funded by the Korea government Ministry of Trade Industry & Energy (N009700001).

4. CONCLUSION In this work, we studied hydrate formation characteristics including hydrate onset time, subcooling temperature, hydrate fraction, and torque changes for pure water, 0.2 wt % PVCap, 20 wt % MEG, 30 wt % MEG, 0.2 wt % PVCap + 20 wt % MEG, and 0.2 wt % + 30 wt % MEG solutions. The average hydrate onset time for pure water was 19.8 min, and the average hydrate fraction was 0.75. The addition of 0.2 wt % PVCap increases the hydrate onset time to 33.8 min; however, the hyradrate fraction was 0.65. Distinct reduction of hydrate fraction was achieved in underinhibited 20 and 30 wt % MEG solutions. The calculated hydrate fraction at the end of the experiment was 0.28 and 0.17, respectively. The high hydrate fraction in pure water and 0.2 wt % PVCap solution induces high torque values, indicating the agglomeration and deposition of hydrate particles; however, the torque changes were negligible during the hydrate formation in 20 and 30 wt % MEG solutions. However, the hydrate onset time for 20 and 30 wt % MEG solutions were 22.8 and 27.3 min, respectively. This indicates that kinetic inhibition performance of the underinhibition systems might be weak to achieve the desired delay time. With the aim of finding alternative methods for hydrate inhibition, we studied the effect of adding PVCap to MEG solutions on the hydrate formation characteristics, which has not been studied thoroughly. It is noted that the addition of 0.2 wt % PVCap to the 20 wt % MEG solution delays the hydrate onset time substantially to 111.0 min and reduces the hydrate fraction below 0.19. Along with the enhanced kinetic hydrate inhibition performance, the torque changes were negligible as well during the hydrate formation process, suggesting the homogeneous dispersion of hydrate particles in the liquid phase. Simple structure II hydrate was confirmed by in situ Raman spectroscopy, while coexisting structures I and II are

■

REFERENCES

(1) Koh, C. A.; Sum, A. K.; Sloan, E. D. State of the Art: Natural Gas Hydrates as a Natural Resource. J. Nat. Gas Sci. Eng. 2012, 8, 132−138. (2) Ohgaki, K.; Takano, K.; Sangawa, H.; Matsubara, T.; Nakano, S. Methane Exploitation by Carbon Dioxide from Gas HydratesPhase Equilibria for CO2-CH4 Mixed Hydrate System. J. Chem. Eng. Jpn. 1996, 29, 478−483. (3) Lee, H.; Seo, Y.; Seo, Y. T.; Moudrakovski, I. L.; Ripmeester, J. A. Recovering Methane from Solid Methane Hydrate with Carbon Dioxide. Angew. Chem., Int. Ed 2003, 42, 5048−5051. (4) Bai, D.; Zhang, X.; Chen, G.; Wang, W. Replacement Mechanism of Methane Hydrate with Carbon Dioxide from Microsecond Molecular Dynamic Simulations. Energy Environ. Sci. 2012, 5, 7033− 7041. (5) Sloan, E. D.; Koh, C. A. Clathrate Hydrates of Natural Gases, 3rd ed.; CRC Press, Taylor & Francis Group: Boca Raton, FL, 2008. (6) Sloan, E. D.; Koh, C. A.; Sum, A. K. Natural Gas Hydrates in Flow Assurance. Elsevier: Amsterdam, 2010. (7) Kelland, M. A. History of the Development of Low Dosage Hydrate Inhibitors. Energy Fuels 2006, 20, 825−847. (8) Daraboina, N.; Linga, P.; Ripmeester, J. A.; Walker, V. K.; Englezos, P. Natural Gas Hydrate Formation and Decomposition in the Presence of Kinetic Inhibitors. 2. Stirred Reactor Experiments. Energy Fuels 2011, 25, 4384−4391. (9) Townson, I.; Walker, V. K.; Ripmeester, J. A.; Englezos, P. Bacterial Inhibition of Methane Clathrate Hydrates Formed in a Stirred Autoclave. Energy Fuels 2012, 26, 7170−7175. 9074

dx.doi.org/10.1021/jp503435t | J. Phys. Chem. B 2014, 118, 9065−9075

The Journal of Physical Chemistry B

Article

(10) Kelland, M. A.; Del Villano, L.; Kommedal, R. Class of Kinetic Hydrate Inhibitors with Good Biodegradability. Energy Fuels 2008, 22, 3143−3149. (11) Kvamme, B.; Kuznetsova, T.; Aasoldsen, K. Molecular Dynamics Simulations for Selection of Kinetic Hydrate Inhibitors. J. Mol. Graphics Modell. 2005, 23, 524−536. (12) Anderson, B. J.; Tester, J. W.; Borshi, G. P.; Trout, B. L. Properties of Inhibitors of Methane Hydrate Formation via Molecular Dynamics Simulations. J. Am. Chem. Soc. 2005, 127, 17852−17862. (13) Anderson, R.; Mozaffar, H.; Tohidi, B. Development of a Crystal Growth Inhibition Based Method for the Evaluation of Kinetic Hydrate Inhibitors. In Proceedings of the 7th International Conference on Gas Hydrates, Edinburgh, Scotland, U.K., July 17−21, 2011. (14) Brustad, S.; Loken, K. P.; Waalmann, J. G. Hydrate Prevention using MEG Instead of MeOH: Impact of Experience from Major Norwegian Developments on Technology Selection for Injection and Recovery of MEG. Offshore Technology Conference, Houston, TX, May 2−5, 2005. (15) Ohno, H.; Strobel, T. A.; Dec, S. F.; Sloan, E. D.; Koh, C. A. Raman Studies of Methane−Ethane Hydrate Metastability. J. Phys. Chem. A 2009, 113, 1711−1716. (16) Ohno, H.; Moudrakovski, I.; Gordienko, R.; Ripmeester, J. A.; Walker, V. K. Structures of Hydrocarbon Hydrates during Formation with and without Inhibitors. J. Phys. Chem. A 2012, 116, 1337−1343. (17) Cha, M.; Shin, K.; Seo, Y.; Shin, J. Y.; Kang, S. P. Catastrophic Growth of Gas Hydrates in the Presence of Kinetic Hydrate Inhibitors. J. Phys. Chem. A 2013, 117, 13988−13995. (18) Moore, J. A. Understanding Kinetic Hydrate Inhibitor and Corrosion Inhibitor Interactions. In Proceedings of the Offshore Technology Conference. Houston, TX, May 4−7, 2009. (19) Joshi, S. V.; Grasso, G. A.; Lafond, P. G.; Rao, I.; Webb, E.; Zerpa, L. E.; Sloan, E. D.; Koh, C. A.; Sum, A. K. Experimental Flowloop Investigations of Gas Hydrate Formation in High Water Cut Systems. Chem. Eng. Sci. 2013, 97, 198−209. (20) Lee, J. D.; Englezos, P. Enhancement of the Performance of Gas Hydrate Kinetic Inhibitors with Polyethylene Oxide. Chem. Eng. Sci. 2005, 60, 5323−5330. (21) Yang, J.; Tohidi, B. Characterization of Inhibition Mechanisms of Kinetic Hydrate Inhibitors using Ultrasonic Test Technique. Chem. Eng. Sci. 2011, 66, 278−283. (22) Hong, S. Y.; Lim, J. I.; Kim, J. H.; Lee, J. D. Kinetic Studies on Methane Hydrate Formation in the Presence of Kinetic Inhibitor via in-situ Raman Spectroscopy. Energy Fuels 2012, 26, 7045−7050. (23) Seo, Y.; Kang, S.-P.; Jang, W. Structure and Composition Analysis of Natural Gas Hydrates: 13C NMR Spectroscopic and Gas Uptake Measurements of Mixed Gas Hydrates. J. Phys. Chem. A 2009, 113, 9641−9649. (24) Ke, W.; Svartaas, T. M.; Abay, H. K. An Experimental Study on sI Hydrate Formation in Presence of Methanol, PVP and PVCap in an Isochoric Cell. In Proceedings of the 7th International Conference on Gas Hydrates, Edinburgh, Scotland, U.K., July 17−21, 2011. (25) Akhfash, M.; Boxall, J. A.; Aman, Z. M.; Johns, M. L.; May, E. F. Hydrate Formation and Particle Distributions in Gas-Water Systems. Chem. Eng. Sci. 2013, 104, 177−188. (26) Bobev, S.; Tait, K. T. Methanol−Inhibitor or Promoter of the Formation of Gas Hydrates from Deuterated ice? Am. Mineral. 2004, 89, 1208−1214. (27) Cha, M.; Shin, K.; Kim, J.; Chang, D.; Seo, Y.; Lee, H.; Kang, S. P. Thermodynamic and Kinetic Hydrate Inhibition Performance of Aqueous Ethylene Glycol Solutions for Natural Gas. Chem. Eng. Sci. 2013, 99, 184−190. (28) Pironon, J.; Grimmer, J.; Teinturier, S.; Guillaume, D.; Dubessy, J. Dissolved Methane in Water: Temperature Effect on Raman Quantification in Fluid Inclusions. J. Geochem. Explor. 2003, 78−79, 111−115. (29) Zerpa, L. E.; Sloan, E. D.; Sum, A. K.; Koh, C. A. Overview of CSMHyK: A transient hydrate formation model. J. Pet. Sci. Eng. 2012, 98−99, 122−129.

(30) Kashchiev, D.; Firoozabadi, A. Induction time in crystallization of gas hydrates. J. Cryst. Growth 2003, 250, 499−515.

9075

dx.doi.org/10.1021/jp503435t | J. Phys. Chem. B 2014, 118, 9065−9075